diff --git a/checkpoints/checkpoint-2100/1_Pooling/config.json b/checkpoints/checkpoint-2100/1_Pooling/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b68441c3f37d8bee79501ca0afe536bf19753928
--- /dev/null
+++ b/checkpoints/checkpoint-2100/1_Pooling/config.json
@@ -0,0 +1,10 @@
+{
+ "word_embedding_dimension": 1024,
+ "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
+}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-2100/README.md b/checkpoints/checkpoint-2100/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..820d158976fbba158e0c611d6c03da16cf8a48c6
--- /dev/null
+++ b/checkpoints/checkpoint-2100/README.md
@@ -0,0 +1,917 @@
+---
+tags:
+- sentence-transformers
+- sentence-similarity
+- feature-extraction
+- generated_from_trainer
+- dataset_size:7552
+- loss:CoSENTLoss
+base_model: intfloat/multilingual-e5-large-instruct
+widget:
+- source_sentence: How are calibration points linked to equipment?
+ sentences:
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+ - "How does a flow computer generate and store reports?\nA flow computer generates\
+ \ daily or hourly reports to provide users with operational data. These reports\
+ \ are stored in the flow computer's memory in an organized format.\n\nReport structure:\n\
+ - Each report includes:\n- Date and time of the data recording.\n- Data recorded\
+ \ from flow computers.\n\nData storage in tables:\nThe reports are saved in two\
+ \ tables:\n1. Main table (Index):\n - Stores the date, time, and flow computer\
+ \ identifier.\n2. Detail table:\n - Stores the measured values associated with\
+ \ the report.\n\nConnection to the Modbus table:\nThe flow computer's reports\
+ \ are linked to a Modbus table. This table contains the names corresponding to\
+ \ each value in the reports, making it easier to interpret the data."
+ - "What is uncertainty?\nUncertainty is a measure of confidence in the precision\
+ \ and reliability of results obtained from equipment or measurement systems. It\
+ \ quantifies the potential error or margin of error in measurements.\n\nTypes\
+ \ of uncertainty:\nThere are two main types of uncertainty:\n1. Uncertainty of\
+ \ magnitudes (variables):\n - Refers to the uncertainty of specific variables,\
+ \ such as temperature or pressure.\n - It is calculated after calibrating a\
+ \ device or obtained from the equipment manufacturer's manual.\n - This uncertainty\
+ \ serves as a starting point for further calculations related to the equipment.\n\
+ \n2. Uncertainty of the measurement system:\n - Refers to the uncertainty calculated\
+ \ for the overall flow measurement.\n - It depends on the uncertainties of\
+ \ the individual variables (magnitudes) and represents the combined margin of\
+ \ error for the entire system.\n\nKey points:\n- The uncertainties of magnitudes\
+ \ (variables) are the foundation for calculating the uncertainty of the measurement\
+ \ system. Think of them as the \"building blocks.\"\n- Do not confuse the two\
+ \ types of uncertainty:\n - **Uncertainty of magnitudes/variables**: Specific\
+ \ to individual variables (e.g., temperature, pressure).\n - **Uncertainty\
+ \ of the measurement system**: Specific to the overall flow measurement.\n\nDatabase\
+ \ storage for uncertainties:\nIn the database, uncertainty calculations are stored\
+ \ in two separate tables:\n1. Uncertainty of magnitudes (variables):\n - Stores\
+ \ the uncertainty values for specific variables (e.g., temperature, pressure).\n\
+ \n2. Uncertainty of the measurement system:\n - Stores the uncertainty values\
+ \ for the overall flow measurement system.\n\nHow to retrieve uncertainty data:\n\
+ - To find the uncertainty of the measurement system, join the measurement systems\
+ \ table with the uncertainty of the measurement system table.\n- To find the uncertainty\
+ \ of a specific variable (magnitude), join the measurement systems table with\
+ \ the uncertainty of magnitudes (variables) table.\n\nImportant note:\nDo not\
+ \ confuse the two types of uncertainty:\n- If the user requests the uncertainty\
+ \ of the measurement system, use the first join (measurement systems table + uncertainty\
+ \ of the measurement system table).\n- If the user requests the uncertainty of\
+ \ a specific variable (magnitude) in a report, use the second join (measurement\
+ \ systems table + uncertainty of magnitudes table)."
+- source_sentence: What is the primary key of the flow computer table?
+ sentences:
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - "How does a flow computer generate and store reports?\nA flow computer generates\
+ \ daily or hourly reports to provide users with operational data. These reports\
+ \ are stored in the flow computer's memory in an organized format.\n\nReport structure:\n\
+ - Each report includes:\n- Date and time of the data recording.\n- Data recorded\
+ \ from flow computers.\n\nData storage in tables:\nThe reports are saved in two\
+ \ tables:\n1. Main table (Index):\n - Stores the date, time, and flow computer\
+ \ identifier.\n2. Detail table:\n - Stores the measured values associated with\
+ \ the report.\n\nConnection to the Modbus table:\nThe flow computer's reports\
+ \ are linked to a Modbus table. This table contains the names corresponding to\
+ \ each value in the reports, making it easier to interpret the data."
+- source_sentence: Can you provide a sample query to test the retrieval of the uncertainty
+ result for the specified tag and date?
+ sentences:
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - "What is uncertainty?\nUncertainty is a measure of confidence in the precision\
+ \ and reliability of results obtained from equipment or measurement systems. It\
+ \ quantifies the potential error or margin of error in measurements.\n\nTypes\
+ \ of uncertainty:\nThere are two main types of uncertainty:\n1. Uncertainty of\
+ \ magnitudes (variables):\n - Refers to the uncertainty of specific variables,\
+ \ such as temperature or pressure.\n - It is calculated after calibrating a\
+ \ device or obtained from the equipment manufacturer's manual.\n - This uncertainty\
+ \ serves as a starting point for further calculations related to the equipment.\n\
+ \n2. Uncertainty of the measurement system:\n - Refers to the uncertainty calculated\
+ \ for the overall flow measurement.\n - It depends on the uncertainties of\
+ \ the individual variables (magnitudes) and represents the combined margin of\
+ \ error for the entire system.\n\nKey points:\n- The uncertainties of magnitudes\
+ \ (variables) are the foundation for calculating the uncertainty of the measurement\
+ \ system. Think of them as the \"building blocks.\"\n- Do not confuse the two\
+ \ types of uncertainty:\n - **Uncertainty of magnitudes/variables**: Specific\
+ \ to individual variables (e.g., temperature, pressure).\n - **Uncertainty\
+ \ of the measurement system**: Specific to the overall flow measurement.\n\nDatabase\
+ \ storage for uncertainties:\nIn the database, uncertainty calculations are stored\
+ \ in two separate tables:\n1. Uncertainty of magnitudes (variables):\n - Stores\
+ \ the uncertainty values for specific variables (e.g., temperature, pressure).\n\
+ \n2. Uncertainty of the measurement system:\n - Stores the uncertainty values\
+ \ for the overall flow measurement system.\n\nHow to retrieve uncertainty data:\n\
+ - To find the uncertainty of the measurement system, join the measurement systems\
+ \ table with the uncertainty of the measurement system table.\n- To find the uncertainty\
+ \ of a specific variable (magnitude), join the measurement systems table with\
+ \ the uncertainty of magnitudes (variables) table.\n\nImportant note:\nDo not\
+ \ confuse the two types of uncertainty:\n- If the user requests the uncertainty\
+ \ of the measurement system, use the first join (measurement systems table + uncertainty\
+ \ of the measurement system table).\n- If the user requests the uncertainty of\
+ \ a specific variable (magnitude) in a report, use the second join (measurement\
+ \ systems table + uncertainty of magnitudes table)."
+- source_sentence: How are the secondary equipment and measurement system related?
+ sentences:
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - 'What do measurement equipment measure?
+
+ Each equipment measures a physical magnitude, also known as a variable. Based
+ on the type of variable they measure, devices are classified into different categories.
+
+
+ Equipment classification:
+
+ - Primary meter: Assigned by default to equipments like orifice plates.
+
+ - Secondary meter: Assigned by default to equipments like transmitters.
+
+ - Tertiary meter: Used for other types of equipments.
+
+
+ Equipment types in the database:
+
+ The database includes a table listing all equipment types. Examples of equipment
+ types are:
+
+ - Differential pressure transmitters
+
+ - RTDs (Resistance Temperature Detectors)
+
+ - Orifice plates
+
+ - Multivariable transmitters
+
+ - Ultrasonic meters
+
+
+ Meteorological checks for equipments:
+
+ Each equipment type is assigned a meteorological check, which can be either:
+
+ - Calibration: To ensure measurement accuracy.
+
+ - Inspection: To verify proper functioning.
+
+
+ Data storage in tables:
+
+ The database also includes a separate table for equipment classifications, which
+ are:
+
+ - Primary meter
+
+ - Secondary meter
+
+ - Tertiary meter
+
+ So, an equipment has equipment types and this types has classifications.'
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+- source_sentence: What is the table structure for secondary equipment?
+ sentences:
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+datasets:
+- Lauther/measuring-embeddings-v3
+pipeline_tag: sentence-similarity
+library_name: sentence-transformers
+---
+
+# SentenceTransformer based on intfloat/multilingual-e5-large-instruct
+
+This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [intfloat/multilingual-e5-large-instruct](https://huggingface.co/intfloat/multilingual-e5-large-instruct) on the [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) 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.
+
+## Model Details
+
+### Model Description
+- **Model Type:** Sentence Transformer
+- **Base model:** [intfloat/multilingual-e5-large-instruct](https://huggingface.co/intfloat/multilingual-e5-large-instruct)
+- **Maximum Sequence Length:** 512 tokens
+- **Output Dimensionality:** 1024 dimensions
+- **Similarity Function:** Cosine Similarity
+- **Training Dataset:**
+ - [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3)
+
+
+
+### Model Sources
+
+- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
+- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
+- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
+
+### Full Model Architecture
+
+```
+SentenceTransformer(
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: XLMRobertaModel
+ (1): Pooling({'word_embedding_dimension': 1024, '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()
+)
+```
+
+## Usage
+
+### Direct Usage (Sentence Transformers)
+
+First install the Sentence Transformers library:
+
+```bash
+pip install -U sentence-transformers
+```
+
+Then you can load this model and run inference.
+```python
+from sentence_transformers import SentenceTransformer
+
+# Download from the 🤗 Hub
+model = SentenceTransformer("sentence_transformers_model_id")
+# Run inference
+sentences = [
+ 'What is the table structure for secondary equipment?',
+ 'How are flow computers and measurement systems related?\nFlow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.\n\nDatabase terminology:\nIn the database, this relationship is referred to as:\n- Meter streams\n- Meter runs\n- Sections\n\nStorage of the relationship:\nThe relationship between a flow computer and its assigned measurement system is stored in a special table.\n\nUser context:\nWhen a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer.',
+ 'What kind of data store an equipment?\nEquipments can capture meteorological data, such as pressure, temperature, and volume (magnitudes). This data is essential for users to perform various calculations.\n\nData storage:\n- The measured values are stored in a special table in the database for magnitudes. This table contains the values of the variables captured by the equipments.\n- These values are **direct measurements** from the fluid (e.g., raw pressure, temperature, or volume readings). **They are not calculated values**, such as uncertainty.\n- The values stored in the variable values table are **different** from variable uncertainty values, which are calculated separately and represent the margin of error.\n\nAccessing the data:\n- Users typically access the data by referring to the readings from the measurement system, not directly from the individual equipments.\n- The readings are stored in a "variable values" table within the database.\n\nLinking variable names:\nIf the user needs to know the name of a variable, they must link the data to another table that stores information about the types of variables.',
+]
+embeddings = model.encode(sentences)
+print(embeddings.shape)
+# [3, 1024]
+
+# Get the similarity scores for the embeddings
+similarities = model.similarity(embeddings, embeddings)
+print(similarities.shape)
+# [3, 3]
+```
+
+
+
+
+
+
+
+
+
+
+
+## Training Details
+
+### Training Dataset
+
+#### measuring-embeddings-v3
+
+* Dataset: [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) at [1b3cbbe](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3/tree/1b3cbbeb70b63338110491cd3de2950fb40b4f87)
+* Size: 7,552 training samples
+* Columns: sentence1, sentence2, and score
+* Approximate statistics based on the first 1000 samples:
+ | | sentence1 | sentence2 | score |
+ |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:----------------------------------------------------------------|
+ | type | string | string | float |
+ | details | - min: 9 tokens
- mean: 15.96 tokens
- max: 40 tokens
| - min: 120 tokens
- mean: 255.56 tokens
- max: 512 tokens
| - min: 0.0
- mean: 0.22
- max: 0.95
|
+* Samples:
+ | sentence1 | sentence2 | score |
+ |:-------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------|
+ | How can I combine the sub-query with the main query to fetch the last uncertainty report? | What do measurement equipment measure?
Each equipment measures a physical magnitude, also known as a variable. Based on the type of variable they measure, devices are classified into different categories.
Equipment classification:
- Primary meter: Assigned by default to equipments like orifice plates.
- Secondary meter: Assigned by default to equipments like transmitters.
- Tertiary meter: Used for other types of equipments.
Equipment types in the database:
The database includes a table listing all equipment types. Examples of equipment types are:
- Differential pressure transmitters
- RTDs (Resistance Temperature Detectors)
- Orifice plates
- Multivariable transmitters
- Ultrasonic meters
Meteorological checks for equipments:
Each equipment type is assigned a meteorological check, which can be either:
- Calibration: To ensure measurement accuracy.
- Inspection: To verify proper functioning.
Data storage in tables:
The database also includes a separate table for equipment classific... | 0.1 |
+ | What is the column name for the calibration date in the calibration table? | How are flow computers and measurement systems related?
Flow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.
Database terminology:
In the database, this relationship is referred to as:
- Meter streams
- Meter runs
- Sections
Storage of the relationship:
The relationship between a flow computer and its assigned measurement system is stored in a special table.
User context:
When a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer. | 0.1 |
+ | What is the name of the table that contains the flow computer tags? | What is equipment calibration?
Calibration is a metrological verification process used to ensure the accuracy of measurement equipment. It is performed periodically, based on intervals set by the company or a regulatory body.
Purpose of calibration:
The calibration process corrects any deviations in how the equipment measures physical magnitudes (variables). This ensures the equipment provides accurate and reliable data.
Calibration cycles:
There are two main calibration cycles:
1. As-found: Represents the equipment's measurement accuracy before any adjustments are made. This cycle is almost always implemented.
2. As-left: Represents the equipment's measurement accuracy after adjustments are made. This cycle is used depending on regulatory requirements.
Calibration uncertainty:
- Uncertainty is included in the results of a calibration.
- Calibration uncertainty refers to the margin of error in the device's measurements, which also affects the uncertainty of the measured variable or ... | 0.05 |
+* Loss: [CoSENTLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
+ ```json
+ {
+ "scale": 20.0,
+ "similarity_fct": "pairwise_cos_sim"
+ }
+ ```
+
+### Evaluation Dataset
+
+#### measuring-embeddings-v3
+
+* Dataset: [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) at [1b3cbbe](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3/tree/1b3cbbeb70b63338110491cd3de2950fb40b4f87)
+* Size: 1,618 evaluation samples
+* Columns: sentence1, sentence2, and score
+* Approximate statistics based on the first 1000 samples:
+ | | sentence1 | sentence2 | score |
+ |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:----------------------------------------------------------------|
+ | type | string | string | float |
+ | details | - min: 9 tokens
- mean: 15.83 tokens
- max: 40 tokens
| - min: 120 tokens
- mean: 250.41 tokens
- max: 512 tokens
| - min: 0.0
- mean: 0.23
- max: 0.95
|
+* Samples:
+ | sentence1 | sentence2 | score |
+ |:--------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------|
+ | Identify any additional tables or columns that might be needed for the query. | How are flow computers and measurement systems related?
Flow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.
Database terminology:
In the database, this relationship is referred to as:
- Meter streams
- Meter runs
- Sections
Storage of the relationship:
The relationship between a flow computer and its assigned measurement system is stored in a special table.
User context:
When a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer. | 0.2 |
+ | What columns in these tables contain the measurement system tag and the flow computer tag? | How does a flow computer generate and store reports?
A flow computer generates daily or hourly reports to provide users with operational data. These reports are stored in the flow computer's memory in an organized format.
Report structure:
- Each report includes:
- Date and time of the data recording.
- Data recorded from flow computers.
Data storage in tables:
The reports are saved in two tables:
1. Main table (Index):
- Stores the date, time, and flow computer identifier.
2. Detail table:
- Stores the measured values associated with the report.
Connection to the Modbus table:
The flow computer's reports are linked to a Modbus table. This table contains the names corresponding to each value in the reports, making it easier to interpret the data. | 0.1 |
+ | Identify the column that stores the calibration number. | What kind of data store an equipment?
Equipments can capture meteorological data, such as pressure, temperature, and volume (magnitudes). This data is essential for users to perform various calculations.
Data storage:
- The measured values are stored in a special table in the database for magnitudes. This table contains the values of the variables captured by the equipments.
- These values are **direct measurements** from the fluid (e.g., raw pressure, temperature, or volume readings). **They are not calculated values**, such as uncertainty.
- The values stored in the variable values table are **different** from variable uncertainty values, which are calculated separately and represent the margin of error.
Accessing the data:
- Users typically access the data by referring to the readings from the measurement system, not directly from the individual equipments.
- The readings are stored in a "variable values" table within the database.
Linking variable names:
If the user needs to kno... | 0.1 |
+* Loss: [CoSENTLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
+ ```json
+ {
+ "scale": 20.0,
+ "similarity_fct": "pairwise_cos_sim"
+ }
+ ```
+
+### Training Hyperparameters
+#### Non-Default Hyperparameters
+
+- `eval_strategy`: steps
+- `per_device_train_batch_size`: 7
+- `per_device_eval_batch_size`: 7
+- `gradient_accumulation_steps`: 4
+- `learning_rate`: 3e-05
+- `num_train_epochs`: 20
+- `warmup_ratio`: 0.1
+
+#### All Hyperparameters
+Click to expand
+
+- `overwrite_output_dir`: False
+- `do_predict`: False
+- `eval_strategy`: steps
+- `prediction_loss_only`: True
+- `per_device_train_batch_size`: 7
+- `per_device_eval_batch_size`: 7
+- `per_gpu_train_batch_size`: None
+- `per_gpu_eval_batch_size`: None
+- `gradient_accumulation_steps`: 4
+- `eval_accumulation_steps`: None
+- `torch_empty_cache_steps`: None
+- `learning_rate`: 3e-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`: 20
+- `max_steps`: -1
+- `lr_scheduler_type`: linear
+- `lr_scheduler_kwargs`: {}
+- `warmup_ratio`: 0.1
+- `warmup_steps`: 0
+- `log_level`: passive
+- `log_level_replica`: warning
+- `log_on_each_node`: True
+- `logging_nan_inf_filter`: True
+- `save_safetensors`: True
+- `save_on_each_node`: False
+- `save_only_model`: False
+- `restore_callback_states_from_checkpoint`: False
+- `no_cuda`: False
+- `use_cpu`: False
+- `use_mps_device`: False
+- `seed`: 42
+- `data_seed`: None
+- `jit_mode_eval`: False
+- `use_ipex`: False
+- `bf16`: False
+- `fp16`: False
+- `fp16_opt_level`: O1
+- `half_precision_backend`: auto
+- `bf16_full_eval`: False
+- `fp16_full_eval`: False
+- `tf32`: None
+- `local_rank`: 0
+- `ddp_backend`: None
+- `tpu_num_cores`: None
+- `tpu_metrics_debug`: False
+- `debug`: []
+- `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`: False
+- `ignore_data_skip`: False
+- `fsdp`: []
+- `fsdp_min_num_params`: 0
+- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
+- `fsdp_transformer_layer_cls_to_wrap`: None
+- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
+- `deepspeed`: None
+- `label_smoothing_factor`: 0.0
+- `optim`: adamw_torch
+- `optim_args`: None
+- `adafactor`: False
+- `group_by_length`: False
+- `length_column_name`: length
+- `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`: False
+- `resume_from_checkpoint`: None
+- `hub_model_id`: None
+- `hub_strategy`: every_save
+- `hub_private_repo`: None
+- `hub_always_push`: False
+- `gradient_checkpointing`: False
+- `gradient_checkpointing_kwargs`: None
+- `include_inputs_for_metrics`: False
+- `include_for_metrics`: []
+- `eval_do_concat_batches`: True
+- `fp16_backend`: auto
+- `push_to_hub_model_id`: None
+- `push_to_hub_organization`: None
+- `mp_parameters`:
+- `auto_find_batch_size`: False
+- `full_determinism`: False
+- `torchdynamo`: None
+- `ray_scope`: last
+- `ddp_timeout`: 1800
+- `torch_compile`: False
+- `torch_compile_backend`: None
+- `torch_compile_mode`: None
+- `dispatch_batches`: None
+- `split_batches`: None
+- `include_tokens_per_second`: False
+- `include_num_input_tokens_seen`: False
+- `neftune_noise_alpha`: None
+- `optim_target_modules`: None
+- `batch_eval_metrics`: False
+- `eval_on_start`: False
+- `use_liger_kernel`: False
+- `eval_use_gather_object`: False
+- `average_tokens_across_devices`: False
+- `prompts`: None
+- `batch_sampler`: batch_sampler
+- `multi_dataset_batch_sampler`: proportional
+
+Click to expand
+
+| Epoch | Step | Training Loss | Validation Loss |
+|:------:|:----:|:-------------:|:---------------:|
+| 3.9379 | 1060 | 8.5934 | - |
+| 3.9750 | 1070 | 8.006 | - |
+| 4.0148 | 1080 | 9.0081 | - |
+| 4.0519 | 1090 | 8.6706 | - |
+| 4.0890 | 1100 | 9.6146 | - |
+| 4.1260 | 1110 | 9.225 | - |
+| 4.1631 | 1120 | 8.7522 | - |
+| 4.2002 | 1130 | 9.0221 | - |
+| 4.2373 | 1140 | 9.6458 | - |
+| 4.2743 | 1150 | 8.7692 | - |
+| 4.3114 | 1160 | 9.2874 | - |
+| 4.3485 | 1170 | 8.9276 | - |
+| 4.3855 | 1180 | 8.7444 | - |
+| 4.4226 | 1190 | 8.7265 | - |
+| 4.4597 | 1200 | 8.7642 | 2.6471 |
+| 4.4968 | 1210 | 8.8917 | - |
+| 4.5338 | 1220 | 9.2155 | - |
+| 4.5709 | 1230 | 8.6101 | - |
+| 4.6080 | 1240 | 8.9904 | - |
+| 4.6450 | 1250 | 9.3272 | - |
+| 4.6821 | 1260 | 7.9367 | - |
+| 4.7192 | 1270 | 8.5891 | - |
+| 4.7563 | 1280 | 8.6286 | - |
+| 4.7933 | 1290 | 7.9982 | - |
+| 4.8304 | 1300 | 7.5587 | - |
+| 4.8675 | 1310 | 7.9405 | - |
+| 4.9045 | 1320 | 9.7092 | - |
+| 4.9416 | 1330 | 8.1475 | - |
+| 4.9787 | 1340 | 9.3603 | - |
+| 5.0148 | 1350 | 7.6621 | 2.8309 |
+| 5.0519 | 1360 | 9.2301 | - |
+| 5.0890 | 1370 | 9.7789 | - |
+| 5.1260 | 1380 | 9.5359 | - |
+| 5.1631 | 1390 | 10.8065 | - |
+| 5.2002 | 1400 | 10.0149 | - |
+| 5.2373 | 1410 | 10.2582 | - |
+| 5.2743 | 1420 | 10.16 | - |
+| 5.3114 | 1430 | 10.0763 | - |
+| 5.3485 | 1440 | 9.5737 | - |
+| 5.3855 | 1450 | 10.4816 | - |
+| 5.4226 | 1460 | 8.6687 | - |
+| 5.4597 | 1470 | 8.4066 | - |
+| 5.4968 | 1480 | 9.386 | - |
+| 5.5338 | 1490 | 8.3911 | - |
+| 5.5709 | 1500 | 8.8025 | 2.5408 |
+| 5.6080 | 1510 | 8.7939 | - |
+| 5.6450 | 1520 | 9.0903 | - |
+| 5.6821 | 1530 | 8.9878 | - |
+| 5.7192 | 1540 | 8.8642 | - |
+| 5.7563 | 1550 | 8.8625 | - |
+| 5.7933 | 1560 | 8.4105 | - |
+| 5.8304 | 1570 | 9.0163 | - |
+| 5.8675 | 1580 | 8.8947 | - |
+| 5.9045 | 1590 | 8.5647 | - |
+| 5.9416 | 1600 | 7.7047 | - |
+| 5.9787 | 1610 | 8.1484 | - |
+| 6.0148 | 1620 | 8.4079 | - |
+| 6.0519 | 1630 | 8.5027 | - |
+| 6.0890 | 1640 | 8.1805 | - |
+| 6.1260 | 1650 | 8.4519 | 2.5901 |
+| 6.1631 | 1660 | 9.062 | - |
+| 6.2002 | 1670 | 8.8499 | - |
+| 6.2373 | 1680 | 8.6576 | - |
+| 6.2743 | 1690 | 8.4652 | - |
+| 6.3114 | 1700 | 9.0782 | - |
+| 6.3485 | 1710 | 8.1532 | - |
+| 6.3855 | 1720 | 8.5185 | - |
+| 6.4226 | 1730 | 9.5908 | - |
+| 6.4597 | 1740 | 8.4188 | - |
+| 6.4968 | 1750 | 8.1885 | - |
+| 6.5338 | 1760 | 8.7666 | - |
+| 6.5709 | 1770 | 8.6105 | - |
+| 6.6080 | 1780 | 8.664 | - |
+| 6.6450 | 1790 | 8.5294 | - |
+| 6.6821 | 1800 | 9.1857 | 2.4974 |
+| 6.7192 | 1810 | 8.7053 | - |
+| 6.7563 | 1820 | 8.1428 | - |
+| 6.7933 | 1830 | 8.4988 | - |
+| 6.8304 | 1840 | 8.4147 | - |
+| 6.8675 | 1850 | 9.069 | - |
+| 6.9045 | 1860 | 8.4405 | - |
+| 6.9416 | 1870 | 9.2157 | - |
+| 6.9787 | 1880 | 9.5492 | - |
+| 7.0148 | 1890 | 8.1325 | - |
+| 7.0519 | 1900 | 8.324 | - |
+| 7.0890 | 1910 | 7.7097 | - |
+| 7.1260 | 1920 | 8.0982 | - |
+| 7.1631 | 1930 | 7.7669 | - |
+| 7.2002 | 1940 | 7.809 | - |
+| 7.2373 | 1950 | 7.9729 | 2.6108 |
+| 7.2743 | 1960 | 8.2125 | - |
+| 7.3114 | 1970 | 7.7403 | - |
+| 7.3485 | 1980 | 7.5494 | - |
+| 7.3855 | 1990 | 8.2821 | - |
+| 7.4226 | 2000 | 8.1644 | - |
+| 7.4597 | 2010 | 8.1664 | - |
+| 7.4968 | 2020 | 8.5876 | - |
+| 7.5338 | 2030 | 8.2753 | - |
+| 7.5709 | 2040 | 9.2057 | - |
+| 7.6080 | 2050 | 8.0052 | - |
+| 7.6450 | 2060 | 8.4954 | - |
+| 7.6821 | 2070 | 8.0325 | - |
+| 7.7192 | 2080 | 8.2934 | - |
+| 7.7563 | 2090 | 9.4019 | - |
+| 7.7933 | 2100 | 8.874 | 2.4529 |
+
+",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "cls_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "eos_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "mask_token": {
+ "content": "",
+ "lstrip": true,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "pad_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "sep_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "unk_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ }
+}
diff --git a/checkpoints/checkpoint-2100/tokenizer.json b/checkpoints/checkpoint-2100/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ed19c669b24d8e44e6e4c98b8ef4bb83d8cab46
--- /dev/null
+++ b/checkpoints/checkpoint-2100/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de6f09c3f9b891e5b98dd3af9463dcab5a97d5265e288271395324a0577e6c05
+size 133
diff --git a/checkpoints/checkpoint-2100/tokenizer_config.json b/checkpoints/checkpoint-2100/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0df9458a41642f96287ae84a989281a3bdcebacf
--- /dev/null
+++ b/checkpoints/checkpoint-2100/tokenizer_config.json
@@ -0,0 +1,56 @@
+{
+ "added_tokens_decoder": {
+ "0": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "1": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "2": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "3": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "250001": {
+ "content": "",
+ "lstrip": true,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ }
+ },
+ "additional_special_tokens": [],
+ "bos_token": "",
+ "clean_up_tokenization_spaces": true,
+ "cls_token": "",
+ "eos_token": "",
+ "extra_special_tokens": {},
+ "mask_token": "",
+ "model_max_length": 512,
+ "pad_token": "",
+ "sep_token": "",
+ "tokenizer_class": "XLMRobertaTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoints/checkpoint-2100/trainer_state.json b/checkpoints/checkpoint-2100/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e2e8d0304bf21b74089462e3d8ab2cabf7b6bf4
--- /dev/null
+++ b/checkpoints/checkpoint-2100/trainer_state.json
@@ -0,0 +1,1615 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 7.793327154772938,
+ "eval_steps": 150,
+ "global_step": 2100,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.03707136237256719,
+ "grad_norm": 53.36411666870117,
+ "learning_rate": 1.1152416356877324e-06,
+ "loss": 10.8165,
+ "step": 10
+ },
+ {
+ "epoch": 0.07414272474513438,
+ "grad_norm": 61.33858871459961,
+ "learning_rate": 2.2304832713754648e-06,
+ "loss": 10.8463,
+ "step": 20
+ },
+ {
+ "epoch": 0.11121408711770157,
+ "grad_norm": 67.16305541992188,
+ "learning_rate": 3.345724907063197e-06,
+ "loss": 10.821,
+ "step": 30
+ },
+ {
+ "epoch": 0.14828544949026876,
+ "grad_norm": 62.08958053588867,
+ "learning_rate": 4.4609665427509296e-06,
+ "loss": 10.4704,
+ "step": 40
+ },
+ {
+ "epoch": 0.18535681186283595,
+ "grad_norm": 53.839927673339844,
+ "learning_rate": 5.576208178438661e-06,
+ "loss": 10.8568,
+ "step": 50
+ },
+ {
+ "epoch": 0.22242817423540315,
+ "grad_norm": 57.02272033691406,
+ "learning_rate": 6.691449814126394e-06,
+ "loss": 10.3005,
+ "step": 60
+ },
+ {
+ "epoch": 0.2594995366079703,
+ "grad_norm": 88.16449737548828,
+ "learning_rate": 7.806691449814127e-06,
+ "loss": 9.6009,
+ "step": 70
+ },
+ {
+ "epoch": 0.2965708989805375,
+ "grad_norm": 84.35252380371094,
+ "learning_rate": 8.921933085501859e-06,
+ "loss": 9.7448,
+ "step": 80
+ },
+ {
+ "epoch": 0.3336422613531047,
+ "grad_norm": 77.93982696533203,
+ "learning_rate": 1.0037174721189593e-05,
+ "loss": 9.5922,
+ "step": 90
+ },
+ {
+ "epoch": 0.3707136237256719,
+ "grad_norm": 56.669166564941406,
+ "learning_rate": 1.1152416356877323e-05,
+ "loss": 10.0919,
+ "step": 100
+ },
+ {
+ "epoch": 0.4077849860982391,
+ "grad_norm": 56.652488708496094,
+ "learning_rate": 1.2267657992565056e-05,
+ "loss": 10.2733,
+ "step": 110
+ },
+ {
+ "epoch": 0.4448563484708063,
+ "grad_norm": 63.472042083740234,
+ "learning_rate": 1.3382899628252788e-05,
+ "loss": 9.4547,
+ "step": 120
+ },
+ {
+ "epoch": 0.4819277108433735,
+ "grad_norm": 78.0159912109375,
+ "learning_rate": 1.4498141263940521e-05,
+ "loss": 9.1738,
+ "step": 130
+ },
+ {
+ "epoch": 0.5189990732159406,
+ "grad_norm": 68.1931381225586,
+ "learning_rate": 1.5613382899628255e-05,
+ "loss": 10.0876,
+ "step": 140
+ },
+ {
+ "epoch": 0.5560704355885079,
+ "grad_norm": 49.315216064453125,
+ "learning_rate": 1.6728624535315986e-05,
+ "loss": 10.3721,
+ "step": 150
+ },
+ {
+ "epoch": 0.5560704355885079,
+ "eval_loss": 2.559400796890259,
+ "eval_runtime": 34.304,
+ "eval_samples_per_second": 47.166,
+ "eval_steps_per_second": 6.763,
+ "step": 150
+ },
+ {
+ "epoch": 0.593141797961075,
+ "grad_norm": 34.1899528503418,
+ "learning_rate": 1.7843866171003718e-05,
+ "loss": 10.4711,
+ "step": 160
+ },
+ {
+ "epoch": 0.6302131603336423,
+ "grad_norm": 37.074134826660156,
+ "learning_rate": 1.895910780669145e-05,
+ "loss": 9.9383,
+ "step": 170
+ },
+ {
+ "epoch": 0.6672845227062094,
+ "grad_norm": 49.323158264160156,
+ "learning_rate": 2.0074349442379185e-05,
+ "loss": 10.0448,
+ "step": 180
+ },
+ {
+ "epoch": 0.7043558850787767,
+ "grad_norm": 53.07639694213867,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 9.8467,
+ "step": 190
+ },
+ {
+ "epoch": 0.7414272474513438,
+ "grad_norm": 38.345916748046875,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 9.9434,
+ "step": 200
+ },
+ {
+ "epoch": 0.7784986098239111,
+ "grad_norm": 47.55274963378906,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 9.7912,
+ "step": 210
+ },
+ {
+ "epoch": 0.8155699721964782,
+ "grad_norm": 59.17176055908203,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 9.7905,
+ "step": 220
+ },
+ {
+ "epoch": 0.8526413345690455,
+ "grad_norm": 52.57457733154297,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 9.9412,
+ "step": 230
+ },
+ {
+ "epoch": 0.8897126969416126,
+ "grad_norm": 42.251304626464844,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 10.0084,
+ "step": 240
+ },
+ {
+ "epoch": 0.9267840593141798,
+ "grad_norm": 46.63966369628906,
+ "learning_rate": 2.788104089219331e-05,
+ "loss": 10.0139,
+ "step": 250
+ },
+ {
+ "epoch": 0.963855421686747,
+ "grad_norm": 27.772289276123047,
+ "learning_rate": 2.8996282527881043e-05,
+ "loss": 9.3794,
+ "step": 260
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 46.3179931640625,
+ "learning_rate": 2.9987608426270136e-05,
+ "loss": 9.5747,
+ "step": 270
+ },
+ {
+ "epoch": 1.0370713623725671,
+ "grad_norm": 45.746551513671875,
+ "learning_rate": 2.98636926889715e-05,
+ "loss": 9.6763,
+ "step": 280
+ },
+ {
+ "epoch": 1.0741427247451343,
+ "grad_norm": 49.50416564941406,
+ "learning_rate": 2.9739776951672864e-05,
+ "loss": 10.4789,
+ "step": 290
+ },
+ {
+ "epoch": 1.1112140871177016,
+ "grad_norm": 43.72733688354492,
+ "learning_rate": 2.9615861214374226e-05,
+ "loss": 9.7244,
+ "step": 300
+ },
+ {
+ "epoch": 1.1112140871177016,
+ "eval_loss": 2.6879734992980957,
+ "eval_runtime": 34.2522,
+ "eval_samples_per_second": 47.238,
+ "eval_steps_per_second": 6.773,
+ "step": 300
+ },
+ {
+ "epoch": 1.1482854494902688,
+ "grad_norm": 48.68376541137695,
+ "learning_rate": 2.9491945477075588e-05,
+ "loss": 9.9421,
+ "step": 310
+ },
+ {
+ "epoch": 1.185356811862836,
+ "grad_norm": 44.014747619628906,
+ "learning_rate": 2.9368029739776953e-05,
+ "loss": 9.5022,
+ "step": 320
+ },
+ {
+ "epoch": 1.222428174235403,
+ "grad_norm": 49.858642578125,
+ "learning_rate": 2.9244114002478315e-05,
+ "loss": 9.6801,
+ "step": 330
+ },
+ {
+ "epoch": 1.2594995366079704,
+ "grad_norm": 49.97459411621094,
+ "learning_rate": 2.9120198265179677e-05,
+ "loss": 9.25,
+ "step": 340
+ },
+ {
+ "epoch": 1.2965708989805376,
+ "grad_norm": 40.38698196411133,
+ "learning_rate": 2.8996282527881043e-05,
+ "loss": 10.2879,
+ "step": 350
+ },
+ {
+ "epoch": 1.3336422613531047,
+ "grad_norm": 29.797264099121094,
+ "learning_rate": 2.8872366790582405e-05,
+ "loss": 9.97,
+ "step": 360
+ },
+ {
+ "epoch": 1.3707136237256718,
+ "grad_norm": 43.83412551879883,
+ "learning_rate": 2.874845105328377e-05,
+ "loss": 9.3751,
+ "step": 370
+ },
+ {
+ "epoch": 1.407784986098239,
+ "grad_norm": 42.754154205322266,
+ "learning_rate": 2.8624535315985132e-05,
+ "loss": 9.8947,
+ "step": 380
+ },
+ {
+ "epoch": 1.4448563484708064,
+ "grad_norm": 50.08184051513672,
+ "learning_rate": 2.8500619578686494e-05,
+ "loss": 9.4987,
+ "step": 390
+ },
+ {
+ "epoch": 1.4819277108433735,
+ "grad_norm": 41.81029510498047,
+ "learning_rate": 2.8376703841387856e-05,
+ "loss": 10.1515,
+ "step": 400
+ },
+ {
+ "epoch": 1.5189990732159406,
+ "grad_norm": 44.03816223144531,
+ "learning_rate": 2.8252788104089218e-05,
+ "loss": 10.0098,
+ "step": 410
+ },
+ {
+ "epoch": 1.556070435588508,
+ "grad_norm": 42.871856689453125,
+ "learning_rate": 2.8128872366790583e-05,
+ "loss": 9.5566,
+ "step": 420
+ },
+ {
+ "epoch": 1.5931417979610751,
+ "grad_norm": 44.18041229248047,
+ "learning_rate": 2.8004956629491945e-05,
+ "loss": 10.1925,
+ "step": 430
+ },
+ {
+ "epoch": 1.6302131603336423,
+ "grad_norm": 30.012725830078125,
+ "learning_rate": 2.788104089219331e-05,
+ "loss": 9.7408,
+ "step": 440
+ },
+ {
+ "epoch": 1.6672845227062094,
+ "grad_norm": 38.104827880859375,
+ "learning_rate": 2.7757125154894673e-05,
+ "loss": 9.6873,
+ "step": 450
+ },
+ {
+ "epoch": 1.6672845227062094,
+ "eval_loss": 2.5061004161834717,
+ "eval_runtime": 34.2772,
+ "eval_samples_per_second": 47.203,
+ "eval_steps_per_second": 6.768,
+ "step": 450
+ },
+ {
+ "epoch": 1.7043558850787766,
+ "grad_norm": 42.93535232543945,
+ "learning_rate": 2.7633209417596035e-05,
+ "loss": 9.1932,
+ "step": 460
+ },
+ {
+ "epoch": 1.7414272474513437,
+ "grad_norm": 63.935279846191406,
+ "learning_rate": 2.75092936802974e-05,
+ "loss": 9.9204,
+ "step": 470
+ },
+ {
+ "epoch": 1.778498609823911,
+ "grad_norm": 30.906002044677734,
+ "learning_rate": 2.7385377942998762e-05,
+ "loss": 9.5151,
+ "step": 480
+ },
+ {
+ "epoch": 1.8155699721964782,
+ "grad_norm": 48.05679702758789,
+ "learning_rate": 2.7261462205700128e-05,
+ "loss": 9.8418,
+ "step": 490
+ },
+ {
+ "epoch": 1.8526413345690456,
+ "grad_norm": 56.36283874511719,
+ "learning_rate": 2.7137546468401486e-05,
+ "loss": 9.6943,
+ "step": 500
+ },
+ {
+ "epoch": 1.8897126969416127,
+ "grad_norm": 47.22933578491211,
+ "learning_rate": 2.7013630731102848e-05,
+ "loss": 9.6948,
+ "step": 510
+ },
+ {
+ "epoch": 1.9267840593141798,
+ "grad_norm": 36.110008239746094,
+ "learning_rate": 2.6889714993804214e-05,
+ "loss": 9.2274,
+ "step": 520
+ },
+ {
+ "epoch": 1.963855421686747,
+ "grad_norm": 37.13912582397461,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 10.138,
+ "step": 530
+ },
+ {
+ "epoch": 2.0,
+ "grad_norm": 51.37559509277344,
+ "learning_rate": 2.664188351920694e-05,
+ "loss": 8.7934,
+ "step": 540
+ },
+ {
+ "epoch": 2.037071362372567,
+ "grad_norm": 42.32170104980469,
+ "learning_rate": 2.6517967781908303e-05,
+ "loss": 8.8304,
+ "step": 550
+ },
+ {
+ "epoch": 2.0741427247451343,
+ "grad_norm": 42.406219482421875,
+ "learning_rate": 2.6394052044609665e-05,
+ "loss": 8.9593,
+ "step": 560
+ },
+ {
+ "epoch": 2.1112140871177014,
+ "grad_norm": 52.097023010253906,
+ "learning_rate": 2.627013630731103e-05,
+ "loss": 9.0051,
+ "step": 570
+ },
+ {
+ "epoch": 2.1482854494902686,
+ "grad_norm": 83.86861419677734,
+ "learning_rate": 2.6146220570012392e-05,
+ "loss": 9.2829,
+ "step": 580
+ },
+ {
+ "epoch": 2.185356811862836,
+ "grad_norm": 38.229949951171875,
+ "learning_rate": 2.6022304832713758e-05,
+ "loss": 9.785,
+ "step": 590
+ },
+ {
+ "epoch": 2.2224281742354033,
+ "grad_norm": 50.322513580322266,
+ "learning_rate": 2.5898389095415116e-05,
+ "loss": 9.6401,
+ "step": 600
+ },
+ {
+ "epoch": 2.2224281742354033,
+ "eval_loss": 2.448084592819214,
+ "eval_runtime": 34.2525,
+ "eval_samples_per_second": 47.237,
+ "eval_steps_per_second": 6.773,
+ "step": 600
+ },
+ {
+ "epoch": 2.266913809082484,
+ "grad_norm": 39.366676330566406,
+ "learning_rate": 2.5774473358116482e-05,
+ "loss": 8.7059,
+ "step": 610
+ },
+ {
+ "epoch": 2.303985171455051,
+ "grad_norm": 41.00213623046875,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 9.1285,
+ "step": 620
+ },
+ {
+ "epoch": 2.341056533827618,
+ "grad_norm": 63.442169189453125,
+ "learning_rate": 2.5526641883519206e-05,
+ "loss": 9.3277,
+ "step": 630
+ },
+ {
+ "epoch": 2.3781278962001853,
+ "grad_norm": 51.54194641113281,
+ "learning_rate": 2.540272614622057e-05,
+ "loss": 9.1577,
+ "step": 640
+ },
+ {
+ "epoch": 2.4151992585727524,
+ "grad_norm": 57.583885192871094,
+ "learning_rate": 2.5278810408921933e-05,
+ "loss": 9.6032,
+ "step": 650
+ },
+ {
+ "epoch": 2.4522706209453196,
+ "grad_norm": 60.13545227050781,
+ "learning_rate": 2.51548946716233e-05,
+ "loss": 8.5451,
+ "step": 660
+ },
+ {
+ "epoch": 2.489341983317887,
+ "grad_norm": 139.85794067382812,
+ "learning_rate": 2.503097893432466e-05,
+ "loss": 8.6601,
+ "step": 670
+ },
+ {
+ "epoch": 2.5264133456904543,
+ "grad_norm": 160.73171997070312,
+ "learning_rate": 2.4907063197026023e-05,
+ "loss": 9.522,
+ "step": 680
+ },
+ {
+ "epoch": 2.5634847080630214,
+ "grad_norm": 190.7194366455078,
+ "learning_rate": 2.4783147459727385e-05,
+ "loss": 10.0619,
+ "step": 690
+ },
+ {
+ "epoch": 2.6005560704355886,
+ "grad_norm": 172.5655517578125,
+ "learning_rate": 2.4659231722428747e-05,
+ "loss": 9.6444,
+ "step": 700
+ },
+ {
+ "epoch": 2.6376274328081557,
+ "grad_norm": 280.9183044433594,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 10.7593,
+ "step": 710
+ },
+ {
+ "epoch": 2.674698795180723,
+ "grad_norm": 118.03450775146484,
+ "learning_rate": 2.4411400247831474e-05,
+ "loss": 10.4432,
+ "step": 720
+ },
+ {
+ "epoch": 2.71177015755329,
+ "grad_norm": 332.662353515625,
+ "learning_rate": 2.428748451053284e-05,
+ "loss": 9.6033,
+ "step": 730
+ },
+ {
+ "epoch": 2.748841519925857,
+ "grad_norm": 129.18788146972656,
+ "learning_rate": 2.41635687732342e-05,
+ "loss": 9.108,
+ "step": 740
+ },
+ {
+ "epoch": 2.7859128822984243,
+ "grad_norm": 350.5856628417969,
+ "learning_rate": 2.4039653035935564e-05,
+ "loss": 9.884,
+ "step": 750
+ },
+ {
+ "epoch": 2.7859128822984243,
+ "eval_loss": 2.599517583847046,
+ "eval_runtime": 34.0592,
+ "eval_samples_per_second": 47.506,
+ "eval_steps_per_second": 6.812,
+ "step": 750
+ },
+ {
+ "epoch": 2.822984244670992,
+ "grad_norm": 285.14862060546875,
+ "learning_rate": 2.391573729863693e-05,
+ "loss": 11.3952,
+ "step": 760
+ },
+ {
+ "epoch": 2.860055607043559,
+ "grad_norm": 185.5862579345703,
+ "learning_rate": 2.379182156133829e-05,
+ "loss": 10.6798,
+ "step": 770
+ },
+ {
+ "epoch": 2.897126969416126,
+ "grad_norm": 77.20765686035156,
+ "learning_rate": 2.3667905824039656e-05,
+ "loss": 9.9795,
+ "step": 780
+ },
+ {
+ "epoch": 2.9341983317886933,
+ "grad_norm": 88.75109100341797,
+ "learning_rate": 2.3543990086741015e-05,
+ "loss": 9.3694,
+ "step": 790
+ },
+ {
+ "epoch": 2.9712696941612604,
+ "grad_norm": 53.416690826416016,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 8.4631,
+ "step": 800
+ },
+ {
+ "epoch": 3.0111214087117704,
+ "grad_norm": 72.12966918945312,
+ "learning_rate": 2.3296158612143742e-05,
+ "loss": 9.3903,
+ "step": 810
+ },
+ {
+ "epoch": 3.0481927710843375,
+ "grad_norm": 159.3909912109375,
+ "learning_rate": 2.3172242874845104e-05,
+ "loss": 8.9608,
+ "step": 820
+ },
+ {
+ "epoch": 3.0852641334569046,
+ "grad_norm": 161.03713989257812,
+ "learning_rate": 2.304832713754647e-05,
+ "loss": 9.623,
+ "step": 830
+ },
+ {
+ "epoch": 3.122335495829472,
+ "grad_norm": 136.50045776367188,
+ "learning_rate": 2.2924411400247832e-05,
+ "loss": 8.5958,
+ "step": 840
+ },
+ {
+ "epoch": 3.159406858202039,
+ "grad_norm": 60.785926818847656,
+ "learning_rate": 2.2800495662949197e-05,
+ "loss": 8.6278,
+ "step": 850
+ },
+ {
+ "epoch": 3.196478220574606,
+ "grad_norm": 115.2509536743164,
+ "learning_rate": 2.267657992565056e-05,
+ "loss": 8.661,
+ "step": 860
+ },
+ {
+ "epoch": 3.233549582947173,
+ "grad_norm": 144.38912963867188,
+ "learning_rate": 2.255266418835192e-05,
+ "loss": 9.5836,
+ "step": 870
+ },
+ {
+ "epoch": 3.2706209453197403,
+ "grad_norm": 239.74520874023438,
+ "learning_rate": 2.2428748451053287e-05,
+ "loss": 8.6221,
+ "step": 880
+ },
+ {
+ "epoch": 3.3076923076923075,
+ "grad_norm": 215.12789916992188,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 9.1995,
+ "step": 890
+ },
+ {
+ "epoch": 3.344763670064875,
+ "grad_norm": 113.77183532714844,
+ "learning_rate": 2.218091697645601e-05,
+ "loss": 8.6909,
+ "step": 900
+ },
+ {
+ "epoch": 3.344763670064875,
+ "eval_loss": 2.4246325492858887,
+ "eval_runtime": 34.1351,
+ "eval_samples_per_second": 47.4,
+ "eval_steps_per_second": 6.797,
+ "step": 900
+ },
+ {
+ "epoch": 3.381835032437442,
+ "grad_norm": 175.02731323242188,
+ "learning_rate": 2.2057001239157373e-05,
+ "loss": 8.7632,
+ "step": 910
+ },
+ {
+ "epoch": 3.4189063948100094,
+ "grad_norm": 142.76580810546875,
+ "learning_rate": 2.1933085501858735e-05,
+ "loss": 8.5604,
+ "step": 920
+ },
+ {
+ "epoch": 3.4559777571825765,
+ "grad_norm": 163.0391387939453,
+ "learning_rate": 2.18091697645601e-05,
+ "loss": 9.3871,
+ "step": 930
+ },
+ {
+ "epoch": 3.4930491195551436,
+ "grad_norm": 386.4294738769531,
+ "learning_rate": 2.1685254027261462e-05,
+ "loss": 9.6445,
+ "step": 940
+ },
+ {
+ "epoch": 3.5301204819277108,
+ "grad_norm": 86.92581939697266,
+ "learning_rate": 2.1561338289962827e-05,
+ "loss": 9.635,
+ "step": 950
+ },
+ {
+ "epoch": 3.567191844300278,
+ "grad_norm": 80.94845581054688,
+ "learning_rate": 2.143742255266419e-05,
+ "loss": 8.8689,
+ "step": 960
+ },
+ {
+ "epoch": 3.6042632066728455,
+ "grad_norm": 69.99708557128906,
+ "learning_rate": 2.1313506815365555e-05,
+ "loss": 9.1411,
+ "step": 970
+ },
+ {
+ "epoch": 3.641334569045412,
+ "grad_norm": 80.53229522705078,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 8.7552,
+ "step": 980
+ },
+ {
+ "epoch": 3.67840593141798,
+ "grad_norm": 53.3475341796875,
+ "learning_rate": 2.1065675340768275e-05,
+ "loss": 8.5697,
+ "step": 990
+ },
+ {
+ "epoch": 3.715477293790547,
+ "grad_norm": 54.19875717163086,
+ "learning_rate": 2.094175960346964e-05,
+ "loss": 8.9008,
+ "step": 1000
+ },
+ {
+ "epoch": 3.752548656163114,
+ "grad_norm": 52.96628952026367,
+ "learning_rate": 2.0817843866171003e-05,
+ "loss": 8.9415,
+ "step": 1010
+ },
+ {
+ "epoch": 3.789620018535681,
+ "grad_norm": 64.08312225341797,
+ "learning_rate": 2.0693928128872368e-05,
+ "loss": 8.1303,
+ "step": 1020
+ },
+ {
+ "epoch": 3.8266913809082483,
+ "grad_norm": 68.22163391113281,
+ "learning_rate": 2.057001239157373e-05,
+ "loss": 8.241,
+ "step": 1030
+ },
+ {
+ "epoch": 3.8637627432808155,
+ "grad_norm": 66.70902252197266,
+ "learning_rate": 2.0446096654275092e-05,
+ "loss": 8.6969,
+ "step": 1040
+ },
+ {
+ "epoch": 3.9008341056533826,
+ "grad_norm": 49.94658279418945,
+ "learning_rate": 2.0322180916976458e-05,
+ "loss": 9.8392,
+ "step": 1050
+ },
+ {
+ "epoch": 3.9008341056533826,
+ "eval_loss": 2.533297300338745,
+ "eval_runtime": 34.1106,
+ "eval_samples_per_second": 47.434,
+ "eval_steps_per_second": 6.801,
+ "step": 1050
+ },
+ {
+ "epoch": 3.93790546802595,
+ "grad_norm": 73.9527816772461,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 8.5934,
+ "step": 1060
+ },
+ {
+ "epoch": 3.974976830398517,
+ "grad_norm": 76.90135955810547,
+ "learning_rate": 2.6703841387856257e-05,
+ "loss": 8.006,
+ "step": 1070
+ },
+ {
+ "epoch": 4.014828544949027,
+ "grad_norm": 85.86236572265625,
+ "learning_rate": 2.664188351920694e-05,
+ "loss": 9.0081,
+ "step": 1080
+ },
+ {
+ "epoch": 4.051899907321594,
+ "grad_norm": 106.57842254638672,
+ "learning_rate": 2.6579925650557622e-05,
+ "loss": 8.6706,
+ "step": 1090
+ },
+ {
+ "epoch": 4.088971269694161,
+ "grad_norm": 68.10344696044922,
+ "learning_rate": 2.6517967781908303e-05,
+ "loss": 9.6146,
+ "step": 1100
+ },
+ {
+ "epoch": 4.126042632066729,
+ "grad_norm": 67.02091217041016,
+ "learning_rate": 2.6456009913258984e-05,
+ "loss": 9.225,
+ "step": 1110
+ },
+ {
+ "epoch": 4.163113994439295,
+ "grad_norm": 67.66644287109375,
+ "learning_rate": 2.6394052044609665e-05,
+ "loss": 8.7522,
+ "step": 1120
+ },
+ {
+ "epoch": 4.200185356811863,
+ "grad_norm": 59.26814651489258,
+ "learning_rate": 2.633209417596035e-05,
+ "loss": 9.0221,
+ "step": 1130
+ },
+ {
+ "epoch": 4.23725671918443,
+ "grad_norm": 88.07310485839844,
+ "learning_rate": 2.627013630731103e-05,
+ "loss": 9.6458,
+ "step": 1140
+ },
+ {
+ "epoch": 4.274328081556997,
+ "grad_norm": 69.33895874023438,
+ "learning_rate": 2.620817843866171e-05,
+ "loss": 8.7692,
+ "step": 1150
+ },
+ {
+ "epoch": 4.311399443929565,
+ "grad_norm": 43.823951721191406,
+ "learning_rate": 2.6146220570012392e-05,
+ "loss": 9.2874,
+ "step": 1160
+ },
+ {
+ "epoch": 4.348470806302132,
+ "grad_norm": 63.590980529785156,
+ "learning_rate": 2.6084262701363077e-05,
+ "loss": 8.9276,
+ "step": 1170
+ },
+ {
+ "epoch": 4.385542168674699,
+ "grad_norm": 43.4519157409668,
+ "learning_rate": 2.6022304832713758e-05,
+ "loss": 8.7444,
+ "step": 1180
+ },
+ {
+ "epoch": 4.422613531047266,
+ "grad_norm": 72.1529312133789,
+ "learning_rate": 2.5960346964064435e-05,
+ "loss": 8.7265,
+ "step": 1190
+ },
+ {
+ "epoch": 4.459684893419833,
+ "grad_norm": 101.67008972167969,
+ "learning_rate": 2.5898389095415116e-05,
+ "loss": 8.7642,
+ "step": 1200
+ },
+ {
+ "epoch": 4.459684893419833,
+ "eval_loss": 2.6471314430236816,
+ "eval_runtime": 34.2723,
+ "eval_samples_per_second": 47.21,
+ "eval_steps_per_second": 6.769,
+ "step": 1200
+ },
+ {
+ "epoch": 4.4967562557924,
+ "grad_norm": 66.03606414794922,
+ "learning_rate": 2.5836431226765797e-05,
+ "loss": 8.8917,
+ "step": 1210
+ },
+ {
+ "epoch": 4.533827618164968,
+ "grad_norm": 60.01662826538086,
+ "learning_rate": 2.5774473358116482e-05,
+ "loss": 9.2155,
+ "step": 1220
+ },
+ {
+ "epoch": 4.570898980537534,
+ "grad_norm": 83.86510467529297,
+ "learning_rate": 2.5712515489467163e-05,
+ "loss": 8.6101,
+ "step": 1230
+ },
+ {
+ "epoch": 4.607970342910102,
+ "grad_norm": 64.35486602783203,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 8.9904,
+ "step": 1240
+ },
+ {
+ "epoch": 4.645041705282669,
+ "grad_norm": 60.44017791748047,
+ "learning_rate": 2.5588599752168525e-05,
+ "loss": 9.3272,
+ "step": 1250
+ },
+ {
+ "epoch": 4.682113067655236,
+ "grad_norm": 56.501102447509766,
+ "learning_rate": 2.5526641883519206e-05,
+ "loss": 7.9367,
+ "step": 1260
+ },
+ {
+ "epoch": 4.719184430027804,
+ "grad_norm": 65.01017761230469,
+ "learning_rate": 2.546468401486989e-05,
+ "loss": 8.5891,
+ "step": 1270
+ },
+ {
+ "epoch": 4.7562557924003706,
+ "grad_norm": 63.42623519897461,
+ "learning_rate": 2.540272614622057e-05,
+ "loss": 8.6286,
+ "step": 1280
+ },
+ {
+ "epoch": 4.793327154772938,
+ "grad_norm": 58.96747589111328,
+ "learning_rate": 2.5340768277571252e-05,
+ "loss": 7.9982,
+ "step": 1290
+ },
+ {
+ "epoch": 4.830398517145505,
+ "grad_norm": 96.77368927001953,
+ "learning_rate": 2.5278810408921933e-05,
+ "loss": 7.5587,
+ "step": 1300
+ },
+ {
+ "epoch": 4.867469879518072,
+ "grad_norm": 85.92269897460938,
+ "learning_rate": 2.5216852540272614e-05,
+ "loss": 7.9405,
+ "step": 1310
+ },
+ {
+ "epoch": 4.904541241890639,
+ "grad_norm": 86.37085723876953,
+ "learning_rate": 2.51548946716233e-05,
+ "loss": 9.7092,
+ "step": 1320
+ },
+ {
+ "epoch": 4.941612604263207,
+ "grad_norm": 65.88274383544922,
+ "learning_rate": 2.509293680297398e-05,
+ "loss": 8.1475,
+ "step": 1330
+ },
+ {
+ "epoch": 4.978683966635774,
+ "grad_norm": 53.252742767333984,
+ "learning_rate": 2.503097893432466e-05,
+ "loss": 9.3603,
+ "step": 1340
+ },
+ {
+ "epoch": 5.014828544949027,
+ "grad_norm": 182.35781860351562,
+ "learning_rate": 2.4969021065675342e-05,
+ "loss": 7.6621,
+ "step": 1350
+ },
+ {
+ "epoch": 5.014828544949027,
+ "eval_loss": 2.8308892250061035,
+ "eval_runtime": 34.1063,
+ "eval_samples_per_second": 47.44,
+ "eval_steps_per_second": 6.802,
+ "step": 1350
+ },
+ {
+ "epoch": 5.051899907321594,
+ "grad_norm": 324.6659240722656,
+ "learning_rate": 2.4907063197026023e-05,
+ "loss": 9.2301,
+ "step": 1360
+ },
+ {
+ "epoch": 5.088971269694161,
+ "grad_norm": 4839.4453125,
+ "learning_rate": 2.4845105328376707e-05,
+ "loss": 9.7789,
+ "step": 1370
+ },
+ {
+ "epoch": 5.126042632066729,
+ "grad_norm": 408588.40625,
+ "learning_rate": 2.4783147459727385e-05,
+ "loss": 9.5359,
+ "step": 1380
+ },
+ {
+ "epoch": 5.163113994439295,
+ "grad_norm": 932.5144653320312,
+ "learning_rate": 2.4721189591078066e-05,
+ "loss": 10.8065,
+ "step": 1390
+ },
+ {
+ "epoch": 5.200185356811863,
+ "grad_norm": 148.0634765625,
+ "learning_rate": 2.4659231722428747e-05,
+ "loss": 10.0149,
+ "step": 1400
+ },
+ {
+ "epoch": 5.23725671918443,
+ "grad_norm": 1214.7279052734375,
+ "learning_rate": 2.459727385377943e-05,
+ "loss": 10.2582,
+ "step": 1410
+ },
+ {
+ "epoch": 5.274328081556997,
+ "grad_norm": 188.14292907714844,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 10.16,
+ "step": 1420
+ },
+ {
+ "epoch": 5.311399443929565,
+ "grad_norm": 388.7134094238281,
+ "learning_rate": 2.4473358116480793e-05,
+ "loss": 10.0763,
+ "step": 1430
+ },
+ {
+ "epoch": 5.348470806302132,
+ "grad_norm": 270.61279296875,
+ "learning_rate": 2.4411400247831474e-05,
+ "loss": 9.5737,
+ "step": 1440
+ },
+ {
+ "epoch": 5.385542168674699,
+ "grad_norm": 1637.854736328125,
+ "learning_rate": 2.4349442379182155e-05,
+ "loss": 10.4816,
+ "step": 1450
+ },
+ {
+ "epoch": 5.422613531047266,
+ "grad_norm": 1118.7012939453125,
+ "learning_rate": 2.428748451053284e-05,
+ "loss": 8.6687,
+ "step": 1460
+ },
+ {
+ "epoch": 5.459684893419833,
+ "grad_norm": 788.9707641601562,
+ "learning_rate": 2.422552664188352e-05,
+ "loss": 8.4066,
+ "step": 1470
+ },
+ {
+ "epoch": 5.4967562557924,
+ "grad_norm": 185.22950744628906,
+ "learning_rate": 2.41635687732342e-05,
+ "loss": 9.386,
+ "step": 1480
+ },
+ {
+ "epoch": 5.533827618164968,
+ "grad_norm": 257.726806640625,
+ "learning_rate": 2.4101610904584883e-05,
+ "loss": 8.3911,
+ "step": 1490
+ },
+ {
+ "epoch": 5.570898980537534,
+ "grad_norm": 867.30908203125,
+ "learning_rate": 2.4039653035935564e-05,
+ "loss": 8.8025,
+ "step": 1500
+ },
+ {
+ "epoch": 5.570898980537534,
+ "eval_loss": 2.5407650470733643,
+ "eval_runtime": 34.0934,
+ "eval_samples_per_second": 47.458,
+ "eval_steps_per_second": 6.805,
+ "step": 1500
+ },
+ {
+ "epoch": 5.607970342910102,
+ "grad_norm": 68.35711669921875,
+ "learning_rate": 2.3977695167286248e-05,
+ "loss": 8.7939,
+ "step": 1510
+ },
+ {
+ "epoch": 5.645041705282669,
+ "grad_norm": 1419.1650390625,
+ "learning_rate": 2.391573729863693e-05,
+ "loss": 9.0903,
+ "step": 1520
+ },
+ {
+ "epoch": 5.682113067655236,
+ "grad_norm": 60.88732147216797,
+ "learning_rate": 2.385377942998761e-05,
+ "loss": 8.9878,
+ "step": 1530
+ },
+ {
+ "epoch": 5.719184430027804,
+ "grad_norm": 61.50148010253906,
+ "learning_rate": 2.379182156133829e-05,
+ "loss": 8.8642,
+ "step": 1540
+ },
+ {
+ "epoch": 5.7562557924003706,
+ "grad_norm": 73.9307861328125,
+ "learning_rate": 2.3729863692688972e-05,
+ "loss": 8.8625,
+ "step": 1550
+ },
+ {
+ "epoch": 5.793327154772938,
+ "grad_norm": 82.28329467773438,
+ "learning_rate": 2.3667905824039656e-05,
+ "loss": 8.4105,
+ "step": 1560
+ },
+ {
+ "epoch": 5.830398517145505,
+ "grad_norm": 83.54141235351562,
+ "learning_rate": 2.3605947955390337e-05,
+ "loss": 9.0163,
+ "step": 1570
+ },
+ {
+ "epoch": 5.867469879518072,
+ "grad_norm": 66.59397888183594,
+ "learning_rate": 2.3543990086741015e-05,
+ "loss": 8.8947,
+ "step": 1580
+ },
+ {
+ "epoch": 5.904541241890639,
+ "grad_norm": 55.8841667175293,
+ "learning_rate": 2.3482032218091696e-05,
+ "loss": 8.5647,
+ "step": 1590
+ },
+ {
+ "epoch": 5.941612604263207,
+ "grad_norm": 74.56892395019531,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 7.7047,
+ "step": 1600
+ },
+ {
+ "epoch": 5.978683966635774,
+ "grad_norm": 106.54405212402344,
+ "learning_rate": 2.335811648079306e-05,
+ "loss": 8.1484,
+ "step": 1610
+ },
+ {
+ "epoch": 6.014828544949027,
+ "grad_norm": 107.8216781616211,
+ "learning_rate": 2.3296158612143742e-05,
+ "loss": 8.4079,
+ "step": 1620
+ },
+ {
+ "epoch": 6.051899907321594,
+ "grad_norm": 100.98133087158203,
+ "learning_rate": 2.3234200743494423e-05,
+ "loss": 8.5027,
+ "step": 1630
+ },
+ {
+ "epoch": 6.088971269694161,
+ "grad_norm": 67.31327819824219,
+ "learning_rate": 2.3172242874845104e-05,
+ "loss": 8.1805,
+ "step": 1640
+ },
+ {
+ "epoch": 6.126042632066729,
+ "grad_norm": 61.10900115966797,
+ "learning_rate": 2.311028500619579e-05,
+ "loss": 8.4519,
+ "step": 1650
+ },
+ {
+ "epoch": 6.126042632066729,
+ "eval_loss": 2.590090036392212,
+ "eval_runtime": 34.1392,
+ "eval_samples_per_second": 47.394,
+ "eval_steps_per_second": 6.796,
+ "step": 1650
+ },
+ {
+ "epoch": 6.163113994439295,
+ "grad_norm": 60.91183090209961,
+ "learning_rate": 2.304832713754647e-05,
+ "loss": 9.062,
+ "step": 1660
+ },
+ {
+ "epoch": 6.200185356811863,
+ "grad_norm": 76.2873306274414,
+ "learning_rate": 2.298636926889715e-05,
+ "loss": 8.8499,
+ "step": 1670
+ },
+ {
+ "epoch": 6.23725671918443,
+ "grad_norm": 74.2361068725586,
+ "learning_rate": 2.2924411400247832e-05,
+ "loss": 8.6576,
+ "step": 1680
+ },
+ {
+ "epoch": 6.274328081556997,
+ "grad_norm": 84.99951934814453,
+ "learning_rate": 2.2862453531598513e-05,
+ "loss": 8.4652,
+ "step": 1690
+ },
+ {
+ "epoch": 6.311399443929565,
+ "grad_norm": 86.40947723388672,
+ "learning_rate": 2.2800495662949197e-05,
+ "loss": 9.0782,
+ "step": 1700
+ },
+ {
+ "epoch": 6.348470806302132,
+ "grad_norm": 75.45568084716797,
+ "learning_rate": 2.2738537794299878e-05,
+ "loss": 8.1532,
+ "step": 1710
+ },
+ {
+ "epoch": 6.385542168674699,
+ "grad_norm": 86.65908813476562,
+ "learning_rate": 2.267657992565056e-05,
+ "loss": 8.5185,
+ "step": 1720
+ },
+ {
+ "epoch": 6.422613531047266,
+ "grad_norm": 62.620758056640625,
+ "learning_rate": 2.261462205700124e-05,
+ "loss": 9.5908,
+ "step": 1730
+ },
+ {
+ "epoch": 6.459684893419833,
+ "grad_norm": 73.63095092773438,
+ "learning_rate": 2.255266418835192e-05,
+ "loss": 8.4188,
+ "step": 1740
+ },
+ {
+ "epoch": 6.4967562557924,
+ "grad_norm": 55.175045013427734,
+ "learning_rate": 2.2490706319702606e-05,
+ "loss": 8.1885,
+ "step": 1750
+ },
+ {
+ "epoch": 6.533827618164968,
+ "grad_norm": 47.647621154785156,
+ "learning_rate": 2.2428748451053287e-05,
+ "loss": 8.7666,
+ "step": 1760
+ },
+ {
+ "epoch": 6.570898980537534,
+ "grad_norm": 191.0419158935547,
+ "learning_rate": 2.2366790582403964e-05,
+ "loss": 8.6105,
+ "step": 1770
+ },
+ {
+ "epoch": 6.607970342910102,
+ "grad_norm": 57.81939697265625,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 8.664,
+ "step": 1780
+ },
+ {
+ "epoch": 6.645041705282669,
+ "grad_norm": 85.04927825927734,
+ "learning_rate": 2.2242874845105326e-05,
+ "loss": 8.5294,
+ "step": 1790
+ },
+ {
+ "epoch": 6.682113067655236,
+ "grad_norm": 55.604122161865234,
+ "learning_rate": 2.218091697645601e-05,
+ "loss": 9.1857,
+ "step": 1800
+ },
+ {
+ "epoch": 6.682113067655236,
+ "eval_loss": 2.497358560562134,
+ "eval_runtime": 34.0441,
+ "eval_samples_per_second": 47.527,
+ "eval_steps_per_second": 6.815,
+ "step": 1800
+ },
+ {
+ "epoch": 6.719184430027804,
+ "grad_norm": 75.21812438964844,
+ "learning_rate": 2.211895910780669e-05,
+ "loss": 8.7053,
+ "step": 1810
+ },
+ {
+ "epoch": 6.7562557924003706,
+ "grad_norm": 85.76676940917969,
+ "learning_rate": 2.2057001239157373e-05,
+ "loss": 8.1428,
+ "step": 1820
+ },
+ {
+ "epoch": 6.793327154772938,
+ "grad_norm": 69.85575866699219,
+ "learning_rate": 2.1995043370508054e-05,
+ "loss": 8.4988,
+ "step": 1830
+ },
+ {
+ "epoch": 6.830398517145505,
+ "grad_norm": 66.85325622558594,
+ "learning_rate": 2.1933085501858735e-05,
+ "loss": 8.4147,
+ "step": 1840
+ },
+ {
+ "epoch": 6.867469879518072,
+ "grad_norm": 60.645294189453125,
+ "learning_rate": 2.187112763320942e-05,
+ "loss": 9.069,
+ "step": 1850
+ },
+ {
+ "epoch": 6.904541241890639,
+ "grad_norm": 77.59921264648438,
+ "learning_rate": 2.18091697645601e-05,
+ "loss": 8.4405,
+ "step": 1860
+ },
+ {
+ "epoch": 6.941612604263207,
+ "grad_norm": 56.27996826171875,
+ "learning_rate": 2.174721189591078e-05,
+ "loss": 9.2157,
+ "step": 1870
+ },
+ {
+ "epoch": 6.978683966635774,
+ "grad_norm": 64.18400573730469,
+ "learning_rate": 2.1685254027261462e-05,
+ "loss": 9.5492,
+ "step": 1880
+ },
+ {
+ "epoch": 7.014828544949027,
+ "grad_norm": 76.78484344482422,
+ "learning_rate": 2.1623296158612143e-05,
+ "loss": 8.1325,
+ "step": 1890
+ },
+ {
+ "epoch": 7.051899907321594,
+ "grad_norm": 77.69871520996094,
+ "learning_rate": 2.1561338289962827e-05,
+ "loss": 8.324,
+ "step": 1900
+ },
+ {
+ "epoch": 7.088971269694161,
+ "grad_norm": 102.43340301513672,
+ "learning_rate": 2.149938042131351e-05,
+ "loss": 7.7097,
+ "step": 1910
+ },
+ {
+ "epoch": 7.126042632066729,
+ "grad_norm": 114.0189208984375,
+ "learning_rate": 2.143742255266419e-05,
+ "loss": 8.0982,
+ "step": 1920
+ },
+ {
+ "epoch": 7.163113994439295,
+ "grad_norm": 66.60838317871094,
+ "learning_rate": 2.137546468401487e-05,
+ "loss": 7.7669,
+ "step": 1930
+ },
+ {
+ "epoch": 7.200185356811863,
+ "grad_norm": 252.72503662109375,
+ "learning_rate": 2.1313506815365555e-05,
+ "loss": 7.809,
+ "step": 1940
+ },
+ {
+ "epoch": 7.23725671918443,
+ "grad_norm": 91.0379409790039,
+ "learning_rate": 2.1251548946716236e-05,
+ "loss": 7.9729,
+ "step": 1950
+ },
+ {
+ "epoch": 7.23725671918443,
+ "eval_loss": 2.610764741897583,
+ "eval_runtime": 34.0774,
+ "eval_samples_per_second": 47.48,
+ "eval_steps_per_second": 6.808,
+ "step": 1950
+ },
+ {
+ "epoch": 7.274328081556997,
+ "grad_norm": 96.09555053710938,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 8.2125,
+ "step": 1960
+ },
+ {
+ "epoch": 7.311399443929565,
+ "grad_norm": 68.51132202148438,
+ "learning_rate": 2.1127633209417594e-05,
+ "loss": 7.7403,
+ "step": 1970
+ },
+ {
+ "epoch": 7.348470806302132,
+ "grad_norm": 94.45308685302734,
+ "learning_rate": 2.1065675340768275e-05,
+ "loss": 7.5494,
+ "step": 1980
+ },
+ {
+ "epoch": 7.385542168674699,
+ "grad_norm": 56.785587310791016,
+ "learning_rate": 2.100371747211896e-05,
+ "loss": 8.2821,
+ "step": 1990
+ },
+ {
+ "epoch": 7.422613531047266,
+ "grad_norm": 91.59173583984375,
+ "learning_rate": 2.094175960346964e-05,
+ "loss": 8.1644,
+ "step": 2000
+ },
+ {
+ "epoch": 7.459684893419833,
+ "grad_norm": 78.57572174072266,
+ "learning_rate": 2.0879801734820322e-05,
+ "loss": 8.1664,
+ "step": 2010
+ },
+ {
+ "epoch": 7.4967562557924,
+ "grad_norm": 71.83062744140625,
+ "learning_rate": 2.0817843866171003e-05,
+ "loss": 8.5876,
+ "step": 2020
+ },
+ {
+ "epoch": 7.533827618164968,
+ "grad_norm": 74.75129699707031,
+ "learning_rate": 2.0755885997521684e-05,
+ "loss": 8.2753,
+ "step": 2030
+ },
+ {
+ "epoch": 7.570898980537534,
+ "grad_norm": 96.26950073242188,
+ "learning_rate": 2.0693928128872368e-05,
+ "loss": 9.2057,
+ "step": 2040
+ },
+ {
+ "epoch": 7.607970342910102,
+ "grad_norm": 96.87144470214844,
+ "learning_rate": 2.063197026022305e-05,
+ "loss": 8.0052,
+ "step": 2050
+ },
+ {
+ "epoch": 7.645041705282669,
+ "grad_norm": 83.44422912597656,
+ "learning_rate": 2.057001239157373e-05,
+ "loss": 8.4954,
+ "step": 2060
+ },
+ {
+ "epoch": 7.682113067655236,
+ "grad_norm": 86.1534194946289,
+ "learning_rate": 2.050805452292441e-05,
+ "loss": 8.0325,
+ "step": 2070
+ },
+ {
+ "epoch": 7.719184430027804,
+ "grad_norm": 89.11744689941406,
+ "learning_rate": 2.0446096654275092e-05,
+ "loss": 8.2934,
+ "step": 2080
+ },
+ {
+ "epoch": 7.7562557924003706,
+ "grad_norm": 60.81666946411133,
+ "learning_rate": 2.0384138785625777e-05,
+ "loss": 9.4019,
+ "step": 2090
+ },
+ {
+ "epoch": 7.793327154772938,
+ "grad_norm": 82.93228912353516,
+ "learning_rate": 2.0322180916976458e-05,
+ "loss": 8.874,
+ "step": 2100
+ },
+ {
+ "epoch": 7.793327154772938,
+ "eval_loss": 2.452894926071167,
+ "eval_runtime": 34.0684,
+ "eval_samples_per_second": 47.493,
+ "eval_steps_per_second": 6.81,
+ "step": 2100
+ }
+ ],
+ "logging_steps": 10,
+ "max_steps": 5380,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 20,
+ "save_steps": 150,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 0.0,
+ "train_batch_size": 7,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-2100/training_args.bin b/checkpoints/checkpoint-2100/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..f1a69c7344d84f80932397ffdc7afef25687d5b4
--- /dev/null
+++ b/checkpoints/checkpoint-2100/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fda94467724e48ac975735648aa63a75558d48b6c993506e918a3476acf80697
+size 129
diff --git a/checkpoints/checkpoint-2550/1_Pooling/config.json b/checkpoints/checkpoint-2550/1_Pooling/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b68441c3f37d8bee79501ca0afe536bf19753928
--- /dev/null
+++ b/checkpoints/checkpoint-2550/1_Pooling/config.json
@@ -0,0 +1,10 @@
+{
+ "word_embedding_dimension": 1024,
+ "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
+}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-2550/README.md b/checkpoints/checkpoint-2550/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..c796fa13d5dc587e703661a5429432cb8d6e5e98
--- /dev/null
+++ b/checkpoints/checkpoint-2550/README.md
@@ -0,0 +1,839 @@
+---
+tags:
+- sentence-transformers
+- sentence-similarity
+- feature-extraction
+- generated_from_trainer
+- dataset_size:7552
+- loss:CoSENTLoss
+base_model: intfloat/multilingual-e5-large-instruct
+widget:
+- source_sentence: How are calibration points linked to equipment?
+ sentences:
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+ - "How does a flow computer generate and store reports?\nA flow computer generates\
+ \ daily or hourly reports to provide users with operational data. These reports\
+ \ are stored in the flow computer's memory in an organized format.\n\nReport structure:\n\
+ - Each report includes:\n- Date and time of the data recording.\n- Data recorded\
+ \ from flow computers.\n\nData storage in tables:\nThe reports are saved in two\
+ \ tables:\n1. Main table (Index):\n - Stores the date, time, and flow computer\
+ \ identifier.\n2. Detail table:\n - Stores the measured values associated with\
+ \ the report.\n\nConnection to the Modbus table:\nThe flow computer's reports\
+ \ are linked to a Modbus table. This table contains the names corresponding to\
+ \ each value in the reports, making it easier to interpret the data."
+ - "What is uncertainty?\nUncertainty is a measure of confidence in the precision\
+ \ and reliability of results obtained from equipment or measurement systems. It\
+ \ quantifies the potential error or margin of error in measurements.\n\nTypes\
+ \ of uncertainty:\nThere are two main types of uncertainty:\n1. Uncertainty of\
+ \ magnitudes (variables):\n - Refers to the uncertainty of specific variables,\
+ \ such as temperature or pressure.\n - It is calculated after calibrating a\
+ \ device or obtained from the equipment manufacturer's manual.\n - This uncertainty\
+ \ serves as a starting point for further calculations related to the equipment.\n\
+ \n2. Uncertainty of the measurement system:\n - Refers to the uncertainty calculated\
+ \ for the overall flow measurement.\n - It depends on the uncertainties of\
+ \ the individual variables (magnitudes) and represents the combined margin of\
+ \ error for the entire system.\n\nKey points:\n- The uncertainties of magnitudes\
+ \ (variables) are the foundation for calculating the uncertainty of the measurement\
+ \ system. Think of them as the \"building blocks.\"\n- Do not confuse the two\
+ \ types of uncertainty:\n - **Uncertainty of magnitudes/variables**: Specific\
+ \ to individual variables (e.g., temperature, pressure).\n - **Uncertainty\
+ \ of the measurement system**: Specific to the overall flow measurement.\n\nDatabase\
+ \ storage for uncertainties:\nIn the database, uncertainty calculations are stored\
+ \ in two separate tables:\n1. Uncertainty of magnitudes (variables):\n - Stores\
+ \ the uncertainty values for specific variables (e.g., temperature, pressure).\n\
+ \n2. Uncertainty of the measurement system:\n - Stores the uncertainty values\
+ \ for the overall flow measurement system.\n\nHow to retrieve uncertainty data:\n\
+ - To find the uncertainty of the measurement system, join the measurement systems\
+ \ table with the uncertainty of the measurement system table.\n- To find the uncertainty\
+ \ of a specific variable (magnitude), join the measurement systems table with\
+ \ the uncertainty of magnitudes (variables) table.\n\nImportant note:\nDo not\
+ \ confuse the two types of uncertainty:\n- If the user requests the uncertainty\
+ \ of the measurement system, use the first join (measurement systems table + uncertainty\
+ \ of the measurement system table).\n- If the user requests the uncertainty of\
+ \ a specific variable (magnitude) in a report, use the second join (measurement\
+ \ systems table + uncertainty of magnitudes table)."
+- source_sentence: What is the primary key of the flow computer table?
+ sentences:
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - "How does a flow computer generate and store reports?\nA flow computer generates\
+ \ daily or hourly reports to provide users with operational data. These reports\
+ \ are stored in the flow computer's memory in an organized format.\n\nReport structure:\n\
+ - Each report includes:\n- Date and time of the data recording.\n- Data recorded\
+ \ from flow computers.\n\nData storage in tables:\nThe reports are saved in two\
+ \ tables:\n1. Main table (Index):\n - Stores the date, time, and flow computer\
+ \ identifier.\n2. Detail table:\n - Stores the measured values associated with\
+ \ the report.\n\nConnection to the Modbus table:\nThe flow computer's reports\
+ \ are linked to a Modbus table. This table contains the names corresponding to\
+ \ each value in the reports, making it easier to interpret the data."
+- source_sentence: Can you provide a sample query to test the retrieval of the uncertainty
+ result for the specified tag and date?
+ sentences:
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - "What is uncertainty?\nUncertainty is a measure of confidence in the precision\
+ \ and reliability of results obtained from equipment or measurement systems. It\
+ \ quantifies the potential error or margin of error in measurements.\n\nTypes\
+ \ of uncertainty:\nThere are two main types of uncertainty:\n1. Uncertainty of\
+ \ magnitudes (variables):\n - Refers to the uncertainty of specific variables,\
+ \ such as temperature or pressure.\n - It is calculated after calibrating a\
+ \ device or obtained from the equipment manufacturer's manual.\n - This uncertainty\
+ \ serves as a starting point for further calculations related to the equipment.\n\
+ \n2. Uncertainty of the measurement system:\n - Refers to the uncertainty calculated\
+ \ for the overall flow measurement.\n - It depends on the uncertainties of\
+ \ the individual variables (magnitudes) and represents the combined margin of\
+ \ error for the entire system.\n\nKey points:\n- The uncertainties of magnitudes\
+ \ (variables) are the foundation for calculating the uncertainty of the measurement\
+ \ system. Think of them as the \"building blocks.\"\n- Do not confuse the two\
+ \ types of uncertainty:\n - **Uncertainty of magnitudes/variables**: Specific\
+ \ to individual variables (e.g., temperature, pressure).\n - **Uncertainty\
+ \ of the measurement system**: Specific to the overall flow measurement.\n\nDatabase\
+ \ storage for uncertainties:\nIn the database, uncertainty calculations are stored\
+ \ in two separate tables:\n1. Uncertainty of magnitudes (variables):\n - Stores\
+ \ the uncertainty values for specific variables (e.g., temperature, pressure).\n\
+ \n2. Uncertainty of the measurement system:\n - Stores the uncertainty values\
+ \ for the overall flow measurement system.\n\nHow to retrieve uncertainty data:\n\
+ - To find the uncertainty of the measurement system, join the measurement systems\
+ \ table with the uncertainty of the measurement system table.\n- To find the uncertainty\
+ \ of a specific variable (magnitude), join the measurement systems table with\
+ \ the uncertainty of magnitudes (variables) table.\n\nImportant note:\nDo not\
+ \ confuse the two types of uncertainty:\n- If the user requests the uncertainty\
+ \ of the measurement system, use the first join (measurement systems table + uncertainty\
+ \ of the measurement system table).\n- If the user requests the uncertainty of\
+ \ a specific variable (magnitude) in a report, use the second join (measurement\
+ \ systems table + uncertainty of magnitudes table)."
+- source_sentence: How are the secondary equipment and measurement system related?
+ sentences:
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - 'What do measurement equipment measure?
+
+ Each equipment measures a physical magnitude, also known as a variable. Based
+ on the type of variable they measure, devices are classified into different categories.
+
+
+ Equipment classification:
+
+ - Primary meter: Assigned by default to equipments like orifice plates.
+
+ - Secondary meter: Assigned by default to equipments like transmitters.
+
+ - Tertiary meter: Used for other types of equipments.
+
+
+ Equipment types in the database:
+
+ The database includes a table listing all equipment types. Examples of equipment
+ types are:
+
+ - Differential pressure transmitters
+
+ - RTDs (Resistance Temperature Detectors)
+
+ - Orifice plates
+
+ - Multivariable transmitters
+
+ - Ultrasonic meters
+
+
+ Meteorological checks for equipments:
+
+ Each equipment type is assigned a meteorological check, which can be either:
+
+ - Calibration: To ensure measurement accuracy.
+
+ - Inspection: To verify proper functioning.
+
+
+ Data storage in tables:
+
+ The database also includes a separate table for equipment classifications, which
+ are:
+
+ - Primary meter
+
+ - Secondary meter
+
+ - Tertiary meter
+
+ So, an equipment has equipment types and this types has classifications.'
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+- source_sentence: What is the table structure for secondary equipment?
+ sentences:
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+datasets:
+- Lauther/measuring-embeddings-v3
+pipeline_tag: sentence-similarity
+library_name: sentence-transformers
+---
+
+# SentenceTransformer based on intfloat/multilingual-e5-large-instruct
+
+This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [intfloat/multilingual-e5-large-instruct](https://huggingface.co/intfloat/multilingual-e5-large-instruct) on the [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) 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.
+
+## Model Details
+
+### Model Description
+- **Model Type:** Sentence Transformer
+- **Base model:** [intfloat/multilingual-e5-large-instruct](https://huggingface.co/intfloat/multilingual-e5-large-instruct)
+- **Maximum Sequence Length:** 512 tokens
+- **Output Dimensionality:** 1024 dimensions
+- **Similarity Function:** Cosine Similarity
+- **Training Dataset:**
+ - [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3)
+
+
+
+### Model Sources
+
+- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
+- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
+- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
+
+### Full Model Architecture
+
+```
+SentenceTransformer(
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: XLMRobertaModel
+ (1): Pooling({'word_embedding_dimension': 1024, '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()
+)
+```
+
+## Usage
+
+### Direct Usage (Sentence Transformers)
+
+First install the Sentence Transformers library:
+
+```bash
+pip install -U sentence-transformers
+```
+
+Then you can load this model and run inference.
+```python
+from sentence_transformers import SentenceTransformer
+
+# Download from the 🤗 Hub
+model = SentenceTransformer("sentence_transformers_model_id")
+# Run inference
+sentences = [
+ 'What is the table structure for secondary equipment?',
+ 'How are flow computers and measurement systems related?\nFlow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.\n\nDatabase terminology:\nIn the database, this relationship is referred to as:\n- Meter streams\n- Meter runs\n- Sections\n\nStorage of the relationship:\nThe relationship between a flow computer and its assigned measurement system is stored in a special table.\n\nUser context:\nWhen a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer.',
+ 'What kind of data store an equipment?\nEquipments can capture meteorological data, such as pressure, temperature, and volume (magnitudes). This data is essential for users to perform various calculations.\n\nData storage:\n- The measured values are stored in a special table in the database for magnitudes. This table contains the values of the variables captured by the equipments.\n- These values are **direct measurements** from the fluid (e.g., raw pressure, temperature, or volume readings). **They are not calculated values**, such as uncertainty.\n- The values stored in the variable values table are **different** from variable uncertainty values, which are calculated separately and represent the margin of error.\n\nAccessing the data:\n- Users typically access the data by referring to the readings from the measurement system, not directly from the individual equipments.\n- The readings are stored in a "variable values" table within the database.\n\nLinking variable names:\nIf the user needs to know the name of a variable, they must link the data to another table that stores information about the types of variables.',
+]
+embeddings = model.encode(sentences)
+print(embeddings.shape)
+# [3, 1024]
+
+# Get the similarity scores for the embeddings
+similarities = model.similarity(embeddings, embeddings)
+print(similarities.shape)
+# [3, 3]
+```
+
+
+
+
+
+
+
+
+
+
+
+## Training Details
+
+### Training Dataset
+
+#### measuring-embeddings-v3
+
+* Dataset: [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) at [1b3cbbe](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3/tree/1b3cbbeb70b63338110491cd3de2950fb40b4f87)
+* Size: 7,552 training samples
+* Columns: sentence1, sentence2, and score
+* Approximate statistics based on the first 1000 samples:
+ | | sentence1 | sentence2 | score |
+ |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:----------------------------------------------------------------|
+ | type | string | string | float |
+ | details | - min: 9 tokens
- mean: 15.96 tokens
- max: 40 tokens
| - min: 120 tokens
- mean: 255.56 tokens
- max: 512 tokens
| - min: 0.0
- mean: 0.22
- max: 0.95
|
+* Samples:
+ | sentence1 | sentence2 | score |
+ |:-------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------|
+ | How can I combine the sub-query with the main query to fetch the last uncertainty report? | What do measurement equipment measure?
Each equipment measures a physical magnitude, also known as a variable. Based on the type of variable they measure, devices are classified into different categories.
Equipment classification:
- Primary meter: Assigned by default to equipments like orifice plates.
- Secondary meter: Assigned by default to equipments like transmitters.
- Tertiary meter: Used for other types of equipments.
Equipment types in the database:
The database includes a table listing all equipment types. Examples of equipment types are:
- Differential pressure transmitters
- RTDs (Resistance Temperature Detectors)
- Orifice plates
- Multivariable transmitters
- Ultrasonic meters
Meteorological checks for equipments:
Each equipment type is assigned a meteorological check, which can be either:
- Calibration: To ensure measurement accuracy.
- Inspection: To verify proper functioning.
Data storage in tables:
The database also includes a separate table for equipment classific... | 0.1 |
+ | What is the column name for the calibration date in the calibration table? | How are flow computers and measurement systems related?
Flow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.
Database terminology:
In the database, this relationship is referred to as:
- Meter streams
- Meter runs
- Sections
Storage of the relationship:
The relationship between a flow computer and its assigned measurement system is stored in a special table.
User context:
When a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer. | 0.1 |
+ | What is the name of the table that contains the flow computer tags? | What is equipment calibration?
Calibration is a metrological verification process used to ensure the accuracy of measurement equipment. It is performed periodically, based on intervals set by the company or a regulatory body.
Purpose of calibration:
The calibration process corrects any deviations in how the equipment measures physical magnitudes (variables). This ensures the equipment provides accurate and reliable data.
Calibration cycles:
There are two main calibration cycles:
1. As-found: Represents the equipment's measurement accuracy before any adjustments are made. This cycle is almost always implemented.
2. As-left: Represents the equipment's measurement accuracy after adjustments are made. This cycle is used depending on regulatory requirements.
Calibration uncertainty:
- Uncertainty is included in the results of a calibration.
- Calibration uncertainty refers to the margin of error in the device's measurements, which also affects the uncertainty of the measured variable or ... | 0.05 |
+* Loss: [CoSENTLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
+ ```json
+ {
+ "scale": 20.0,
+ "similarity_fct": "pairwise_cos_sim"
+ }
+ ```
+
+### Evaluation Dataset
+
+#### measuring-embeddings-v3
+
+* Dataset: [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) at [1b3cbbe](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3/tree/1b3cbbeb70b63338110491cd3de2950fb40b4f87)
+* Size: 1,618 evaluation samples
+* Columns: sentence1, sentence2, and score
+* Approximate statistics based on the first 1000 samples:
+ | | sentence1 | sentence2 | score |
+ |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:----------------------------------------------------------------|
+ | type | string | string | float |
+ | details | - min: 9 tokens
- mean: 15.83 tokens
- max: 40 tokens
| - min: 120 tokens
- mean: 250.41 tokens
- max: 512 tokens
| - min: 0.0
- mean: 0.23
- max: 0.95
|
+* Samples:
+ | sentence1 | sentence2 | score |
+ |:--------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------|
+ | Identify any additional tables or columns that might be needed for the query. | How are flow computers and measurement systems related?
Flow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.
Database terminology:
In the database, this relationship is referred to as:
- Meter streams
- Meter runs
- Sections
Storage of the relationship:
The relationship between a flow computer and its assigned measurement system is stored in a special table.
User context:
When a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer. | 0.2 |
+ | What columns in these tables contain the measurement system tag and the flow computer tag? | How does a flow computer generate and store reports?
A flow computer generates daily or hourly reports to provide users with operational data. These reports are stored in the flow computer's memory in an organized format.
Report structure:
- Each report includes:
- Date and time of the data recording.
- Data recorded from flow computers.
Data storage in tables:
The reports are saved in two tables:
1. Main table (Index):
- Stores the date, time, and flow computer identifier.
2. Detail table:
- Stores the measured values associated with the report.
Connection to the Modbus table:
The flow computer's reports are linked to a Modbus table. This table contains the names corresponding to each value in the reports, making it easier to interpret the data. | 0.1 |
+ | Identify the column that stores the calibration number. | What kind of data store an equipment?
Equipments can capture meteorological data, such as pressure, temperature, and volume (magnitudes). This data is essential for users to perform various calculations.
Data storage:
- The measured values are stored in a special table in the database for magnitudes. This table contains the values of the variables captured by the equipments.
- These values are **direct measurements** from the fluid (e.g., raw pressure, temperature, or volume readings). **They are not calculated values**, such as uncertainty.
- The values stored in the variable values table are **different** from variable uncertainty values, which are calculated separately and represent the margin of error.
Accessing the data:
- Users typically access the data by referring to the readings from the measurement system, not directly from the individual equipments.
- The readings are stored in a "variable values" table within the database.
Linking variable names:
If the user needs to kno... | 0.1 |
+* Loss: [CoSENTLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
+ ```json
+ {
+ "scale": 20.0,
+ "similarity_fct": "pairwise_cos_sim"
+ }
+ ```
+
+### Training Hyperparameters
+#### Non-Default Hyperparameters
+
+- `eval_strategy`: steps
+- `per_device_train_batch_size`: 7
+- `per_device_eval_batch_size`: 7
+- `gradient_accumulation_steps`: 4
+- `learning_rate`: 3e-05
+- `num_train_epochs`: 20
+- `warmup_ratio`: 0.1
+
+#### All Hyperparameters
+Click to expand
+
+- `overwrite_output_dir`: False
+- `do_predict`: False
+- `eval_strategy`: steps
+- `prediction_loss_only`: True
+- `per_device_train_batch_size`: 7
+- `per_device_eval_batch_size`: 7
+- `per_gpu_train_batch_size`: None
+- `per_gpu_eval_batch_size`: None
+- `gradient_accumulation_steps`: 4
+- `eval_accumulation_steps`: None
+- `torch_empty_cache_steps`: None
+- `learning_rate`: 3e-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`: 20
+- `max_steps`: -1
+- `lr_scheduler_type`: linear
+- `lr_scheduler_kwargs`: {}
+- `warmup_ratio`: 0.1
+- `warmup_steps`: 0
+- `log_level`: passive
+- `log_level_replica`: warning
+- `log_on_each_node`: True
+- `logging_nan_inf_filter`: True
+- `save_safetensors`: True
+- `save_on_each_node`: False
+- `save_only_model`: False
+- `restore_callback_states_from_checkpoint`: False
+- `no_cuda`: False
+- `use_cpu`: False
+- `use_mps_device`: False
+- `seed`: 42
+- `data_seed`: None
+- `jit_mode_eval`: False
+- `use_ipex`: False
+- `bf16`: False
+- `fp16`: False
+- `fp16_opt_level`: O1
+- `half_precision_backend`: auto
+- `bf16_full_eval`: False
+- `fp16_full_eval`: False
+- `tf32`: None
+- `local_rank`: 0
+- `ddp_backend`: None
+- `tpu_num_cores`: None
+- `tpu_metrics_debug`: False
+- `debug`: []
+- `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`: False
+- `ignore_data_skip`: False
+- `fsdp`: []
+- `fsdp_min_num_params`: 0
+- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
+- `fsdp_transformer_layer_cls_to_wrap`: None
+- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
+- `deepspeed`: None
+- `label_smoothing_factor`: 0.0
+- `optim`: adamw_torch
+- `optim_args`: None
+- `adafactor`: False
+- `group_by_length`: False
+- `length_column_name`: length
+- `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`: False
+- `resume_from_checkpoint`: None
+- `hub_model_id`: None
+- `hub_strategy`: every_save
+- `hub_private_repo`: None
+- `hub_always_push`: False
+- `gradient_checkpointing`: False
+- `gradient_checkpointing_kwargs`: None
+- `include_inputs_for_metrics`: False
+- `include_for_metrics`: []
+- `eval_do_concat_batches`: True
+- `fp16_backend`: auto
+- `push_to_hub_model_id`: None
+- `push_to_hub_organization`: None
+- `mp_parameters`:
+- `auto_find_batch_size`: False
+- `full_determinism`: False
+- `torchdynamo`: None
+- `ray_scope`: last
+- `ddp_timeout`: 1800
+- `torch_compile`: False
+- `torch_compile_backend`: None
+- `torch_compile_mode`: None
+- `dispatch_batches`: None
+- `split_batches`: None
+- `include_tokens_per_second`: False
+- `include_num_input_tokens_seen`: False
+- `neftune_noise_alpha`: None
+- `optim_target_modules`: None
+- `batch_eval_metrics`: False
+- `eval_on_start`: False
+- `use_liger_kernel`: False
+- `eval_use_gather_object`: False
+- `average_tokens_across_devices`: False
+- `prompts`: None
+- `batch_sampler`: batch_sampler
+- `multi_dataset_batch_sampler`: proportional
+
+",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "cls_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "eos_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "mask_token": {
+ "content": "",
+ "lstrip": true,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "pad_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "sep_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "unk_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ }
+}
diff --git a/checkpoints/checkpoint-2550/tokenizer.json b/checkpoints/checkpoint-2550/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ed19c669b24d8e44e6e4c98b8ef4bb83d8cab46
--- /dev/null
+++ b/checkpoints/checkpoint-2550/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de6f09c3f9b891e5b98dd3af9463dcab5a97d5265e288271395324a0577e6c05
+size 133
diff --git a/checkpoints/checkpoint-2550/tokenizer_config.json b/checkpoints/checkpoint-2550/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0df9458a41642f96287ae84a989281a3bdcebacf
--- /dev/null
+++ b/checkpoints/checkpoint-2550/tokenizer_config.json
@@ -0,0 +1,56 @@
+{
+ "added_tokens_decoder": {
+ "0": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "1": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "2": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "3": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "250001": {
+ "content": "",
+ "lstrip": true,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ }
+ },
+ "additional_special_tokens": [],
+ "bos_token": "",
+ "clean_up_tokenization_spaces": true,
+ "cls_token": "",
+ "eos_token": "",
+ "extra_special_tokens": {},
+ "mask_token": "",
+ "model_max_length": 512,
+ "pad_token": "",
+ "sep_token": "",
+ "tokenizer_class": "XLMRobertaTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoints/checkpoint-2550/trainer_state.json b/checkpoints/checkpoint-2550/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..c65b19c725bee45e302d5edb972592a4cceb6613
--- /dev/null
+++ b/checkpoints/checkpoint-2550/trainer_state.json
@@ -0,0 +1,1954 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 9.478220574606116,
+ "eval_steps": 150,
+ "global_step": 2550,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.03707136237256719,
+ "grad_norm": 53.36411666870117,
+ "learning_rate": 1.1152416356877324e-06,
+ "loss": 10.8165,
+ "step": 10
+ },
+ {
+ "epoch": 0.07414272474513438,
+ "grad_norm": 61.33858871459961,
+ "learning_rate": 2.2304832713754648e-06,
+ "loss": 10.8463,
+ "step": 20
+ },
+ {
+ "epoch": 0.11121408711770157,
+ "grad_norm": 67.16305541992188,
+ "learning_rate": 3.345724907063197e-06,
+ "loss": 10.821,
+ "step": 30
+ },
+ {
+ "epoch": 0.14828544949026876,
+ "grad_norm": 62.08958053588867,
+ "learning_rate": 4.4609665427509296e-06,
+ "loss": 10.4704,
+ "step": 40
+ },
+ {
+ "epoch": 0.18535681186283595,
+ "grad_norm": 53.839927673339844,
+ "learning_rate": 5.576208178438661e-06,
+ "loss": 10.8568,
+ "step": 50
+ },
+ {
+ "epoch": 0.22242817423540315,
+ "grad_norm": 57.02272033691406,
+ "learning_rate": 6.691449814126394e-06,
+ "loss": 10.3005,
+ "step": 60
+ },
+ {
+ "epoch": 0.2594995366079703,
+ "grad_norm": 88.16449737548828,
+ "learning_rate": 7.806691449814127e-06,
+ "loss": 9.6009,
+ "step": 70
+ },
+ {
+ "epoch": 0.2965708989805375,
+ "grad_norm": 84.35252380371094,
+ "learning_rate": 8.921933085501859e-06,
+ "loss": 9.7448,
+ "step": 80
+ },
+ {
+ "epoch": 0.3336422613531047,
+ "grad_norm": 77.93982696533203,
+ "learning_rate": 1.0037174721189593e-05,
+ "loss": 9.5922,
+ "step": 90
+ },
+ {
+ "epoch": 0.3707136237256719,
+ "grad_norm": 56.669166564941406,
+ "learning_rate": 1.1152416356877323e-05,
+ "loss": 10.0919,
+ "step": 100
+ },
+ {
+ "epoch": 0.4077849860982391,
+ "grad_norm": 56.652488708496094,
+ "learning_rate": 1.2267657992565056e-05,
+ "loss": 10.2733,
+ "step": 110
+ },
+ {
+ "epoch": 0.4448563484708063,
+ "grad_norm": 63.472042083740234,
+ "learning_rate": 1.3382899628252788e-05,
+ "loss": 9.4547,
+ "step": 120
+ },
+ {
+ "epoch": 0.4819277108433735,
+ "grad_norm": 78.0159912109375,
+ "learning_rate": 1.4498141263940521e-05,
+ "loss": 9.1738,
+ "step": 130
+ },
+ {
+ "epoch": 0.5189990732159406,
+ "grad_norm": 68.1931381225586,
+ "learning_rate": 1.5613382899628255e-05,
+ "loss": 10.0876,
+ "step": 140
+ },
+ {
+ "epoch": 0.5560704355885079,
+ "grad_norm": 49.315216064453125,
+ "learning_rate": 1.6728624535315986e-05,
+ "loss": 10.3721,
+ "step": 150
+ },
+ {
+ "epoch": 0.5560704355885079,
+ "eval_loss": 2.559400796890259,
+ "eval_runtime": 34.304,
+ "eval_samples_per_second": 47.166,
+ "eval_steps_per_second": 6.763,
+ "step": 150
+ },
+ {
+ "epoch": 0.593141797961075,
+ "grad_norm": 34.1899528503418,
+ "learning_rate": 1.7843866171003718e-05,
+ "loss": 10.4711,
+ "step": 160
+ },
+ {
+ "epoch": 0.6302131603336423,
+ "grad_norm": 37.074134826660156,
+ "learning_rate": 1.895910780669145e-05,
+ "loss": 9.9383,
+ "step": 170
+ },
+ {
+ "epoch": 0.6672845227062094,
+ "grad_norm": 49.323158264160156,
+ "learning_rate": 2.0074349442379185e-05,
+ "loss": 10.0448,
+ "step": 180
+ },
+ {
+ "epoch": 0.7043558850787767,
+ "grad_norm": 53.07639694213867,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 9.8467,
+ "step": 190
+ },
+ {
+ "epoch": 0.7414272474513438,
+ "grad_norm": 38.345916748046875,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 9.9434,
+ "step": 200
+ },
+ {
+ "epoch": 0.7784986098239111,
+ "grad_norm": 47.55274963378906,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 9.7912,
+ "step": 210
+ },
+ {
+ "epoch": 0.8155699721964782,
+ "grad_norm": 59.17176055908203,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 9.7905,
+ "step": 220
+ },
+ {
+ "epoch": 0.8526413345690455,
+ "grad_norm": 52.57457733154297,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 9.9412,
+ "step": 230
+ },
+ {
+ "epoch": 0.8897126969416126,
+ "grad_norm": 42.251304626464844,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 10.0084,
+ "step": 240
+ },
+ {
+ "epoch": 0.9267840593141798,
+ "grad_norm": 46.63966369628906,
+ "learning_rate": 2.788104089219331e-05,
+ "loss": 10.0139,
+ "step": 250
+ },
+ {
+ "epoch": 0.963855421686747,
+ "grad_norm": 27.772289276123047,
+ "learning_rate": 2.8996282527881043e-05,
+ "loss": 9.3794,
+ "step": 260
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 46.3179931640625,
+ "learning_rate": 2.9987608426270136e-05,
+ "loss": 9.5747,
+ "step": 270
+ },
+ {
+ "epoch": 1.0370713623725671,
+ "grad_norm": 45.746551513671875,
+ "learning_rate": 2.98636926889715e-05,
+ "loss": 9.6763,
+ "step": 280
+ },
+ {
+ "epoch": 1.0741427247451343,
+ "grad_norm": 49.50416564941406,
+ "learning_rate": 2.9739776951672864e-05,
+ "loss": 10.4789,
+ "step": 290
+ },
+ {
+ "epoch": 1.1112140871177016,
+ "grad_norm": 43.72733688354492,
+ "learning_rate": 2.9615861214374226e-05,
+ "loss": 9.7244,
+ "step": 300
+ },
+ {
+ "epoch": 1.1112140871177016,
+ "eval_loss": 2.6879734992980957,
+ "eval_runtime": 34.2522,
+ "eval_samples_per_second": 47.238,
+ "eval_steps_per_second": 6.773,
+ "step": 300
+ },
+ {
+ "epoch": 1.1482854494902688,
+ "grad_norm": 48.68376541137695,
+ "learning_rate": 2.9491945477075588e-05,
+ "loss": 9.9421,
+ "step": 310
+ },
+ {
+ "epoch": 1.185356811862836,
+ "grad_norm": 44.014747619628906,
+ "learning_rate": 2.9368029739776953e-05,
+ "loss": 9.5022,
+ "step": 320
+ },
+ {
+ "epoch": 1.222428174235403,
+ "grad_norm": 49.858642578125,
+ "learning_rate": 2.9244114002478315e-05,
+ "loss": 9.6801,
+ "step": 330
+ },
+ {
+ "epoch": 1.2594995366079704,
+ "grad_norm": 49.97459411621094,
+ "learning_rate": 2.9120198265179677e-05,
+ "loss": 9.25,
+ "step": 340
+ },
+ {
+ "epoch": 1.2965708989805376,
+ "grad_norm": 40.38698196411133,
+ "learning_rate": 2.8996282527881043e-05,
+ "loss": 10.2879,
+ "step": 350
+ },
+ {
+ "epoch": 1.3336422613531047,
+ "grad_norm": 29.797264099121094,
+ "learning_rate": 2.8872366790582405e-05,
+ "loss": 9.97,
+ "step": 360
+ },
+ {
+ "epoch": 1.3707136237256718,
+ "grad_norm": 43.83412551879883,
+ "learning_rate": 2.874845105328377e-05,
+ "loss": 9.3751,
+ "step": 370
+ },
+ {
+ "epoch": 1.407784986098239,
+ "grad_norm": 42.754154205322266,
+ "learning_rate": 2.8624535315985132e-05,
+ "loss": 9.8947,
+ "step": 380
+ },
+ {
+ "epoch": 1.4448563484708064,
+ "grad_norm": 50.08184051513672,
+ "learning_rate": 2.8500619578686494e-05,
+ "loss": 9.4987,
+ "step": 390
+ },
+ {
+ "epoch": 1.4819277108433735,
+ "grad_norm": 41.81029510498047,
+ "learning_rate": 2.8376703841387856e-05,
+ "loss": 10.1515,
+ "step": 400
+ },
+ {
+ "epoch": 1.5189990732159406,
+ "grad_norm": 44.03816223144531,
+ "learning_rate": 2.8252788104089218e-05,
+ "loss": 10.0098,
+ "step": 410
+ },
+ {
+ "epoch": 1.556070435588508,
+ "grad_norm": 42.871856689453125,
+ "learning_rate": 2.8128872366790583e-05,
+ "loss": 9.5566,
+ "step": 420
+ },
+ {
+ "epoch": 1.5931417979610751,
+ "grad_norm": 44.18041229248047,
+ "learning_rate": 2.8004956629491945e-05,
+ "loss": 10.1925,
+ "step": 430
+ },
+ {
+ "epoch": 1.6302131603336423,
+ "grad_norm": 30.012725830078125,
+ "learning_rate": 2.788104089219331e-05,
+ "loss": 9.7408,
+ "step": 440
+ },
+ {
+ "epoch": 1.6672845227062094,
+ "grad_norm": 38.104827880859375,
+ "learning_rate": 2.7757125154894673e-05,
+ "loss": 9.6873,
+ "step": 450
+ },
+ {
+ "epoch": 1.6672845227062094,
+ "eval_loss": 2.5061004161834717,
+ "eval_runtime": 34.2772,
+ "eval_samples_per_second": 47.203,
+ "eval_steps_per_second": 6.768,
+ "step": 450
+ },
+ {
+ "epoch": 1.7043558850787766,
+ "grad_norm": 42.93535232543945,
+ "learning_rate": 2.7633209417596035e-05,
+ "loss": 9.1932,
+ "step": 460
+ },
+ {
+ "epoch": 1.7414272474513437,
+ "grad_norm": 63.935279846191406,
+ "learning_rate": 2.75092936802974e-05,
+ "loss": 9.9204,
+ "step": 470
+ },
+ {
+ "epoch": 1.778498609823911,
+ "grad_norm": 30.906002044677734,
+ "learning_rate": 2.7385377942998762e-05,
+ "loss": 9.5151,
+ "step": 480
+ },
+ {
+ "epoch": 1.8155699721964782,
+ "grad_norm": 48.05679702758789,
+ "learning_rate": 2.7261462205700128e-05,
+ "loss": 9.8418,
+ "step": 490
+ },
+ {
+ "epoch": 1.8526413345690456,
+ "grad_norm": 56.36283874511719,
+ "learning_rate": 2.7137546468401486e-05,
+ "loss": 9.6943,
+ "step": 500
+ },
+ {
+ "epoch": 1.8897126969416127,
+ "grad_norm": 47.22933578491211,
+ "learning_rate": 2.7013630731102848e-05,
+ "loss": 9.6948,
+ "step": 510
+ },
+ {
+ "epoch": 1.9267840593141798,
+ "grad_norm": 36.110008239746094,
+ "learning_rate": 2.6889714993804214e-05,
+ "loss": 9.2274,
+ "step": 520
+ },
+ {
+ "epoch": 1.963855421686747,
+ "grad_norm": 37.13912582397461,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 10.138,
+ "step": 530
+ },
+ {
+ "epoch": 2.0,
+ "grad_norm": 51.37559509277344,
+ "learning_rate": 2.664188351920694e-05,
+ "loss": 8.7934,
+ "step": 540
+ },
+ {
+ "epoch": 2.037071362372567,
+ "grad_norm": 42.32170104980469,
+ "learning_rate": 2.6517967781908303e-05,
+ "loss": 8.8304,
+ "step": 550
+ },
+ {
+ "epoch": 2.0741427247451343,
+ "grad_norm": 42.406219482421875,
+ "learning_rate": 2.6394052044609665e-05,
+ "loss": 8.9593,
+ "step": 560
+ },
+ {
+ "epoch": 2.1112140871177014,
+ "grad_norm": 52.097023010253906,
+ "learning_rate": 2.627013630731103e-05,
+ "loss": 9.0051,
+ "step": 570
+ },
+ {
+ "epoch": 2.1482854494902686,
+ "grad_norm": 83.86861419677734,
+ "learning_rate": 2.6146220570012392e-05,
+ "loss": 9.2829,
+ "step": 580
+ },
+ {
+ "epoch": 2.185356811862836,
+ "grad_norm": 38.229949951171875,
+ "learning_rate": 2.6022304832713758e-05,
+ "loss": 9.785,
+ "step": 590
+ },
+ {
+ "epoch": 2.2224281742354033,
+ "grad_norm": 50.322513580322266,
+ "learning_rate": 2.5898389095415116e-05,
+ "loss": 9.6401,
+ "step": 600
+ },
+ {
+ "epoch": 2.2224281742354033,
+ "eval_loss": 2.448084592819214,
+ "eval_runtime": 34.2525,
+ "eval_samples_per_second": 47.237,
+ "eval_steps_per_second": 6.773,
+ "step": 600
+ },
+ {
+ "epoch": 2.266913809082484,
+ "grad_norm": 39.366676330566406,
+ "learning_rate": 2.5774473358116482e-05,
+ "loss": 8.7059,
+ "step": 610
+ },
+ {
+ "epoch": 2.303985171455051,
+ "grad_norm": 41.00213623046875,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 9.1285,
+ "step": 620
+ },
+ {
+ "epoch": 2.341056533827618,
+ "grad_norm": 63.442169189453125,
+ "learning_rate": 2.5526641883519206e-05,
+ "loss": 9.3277,
+ "step": 630
+ },
+ {
+ "epoch": 2.3781278962001853,
+ "grad_norm": 51.54194641113281,
+ "learning_rate": 2.540272614622057e-05,
+ "loss": 9.1577,
+ "step": 640
+ },
+ {
+ "epoch": 2.4151992585727524,
+ "grad_norm": 57.583885192871094,
+ "learning_rate": 2.5278810408921933e-05,
+ "loss": 9.6032,
+ "step": 650
+ },
+ {
+ "epoch": 2.4522706209453196,
+ "grad_norm": 60.13545227050781,
+ "learning_rate": 2.51548946716233e-05,
+ "loss": 8.5451,
+ "step": 660
+ },
+ {
+ "epoch": 2.489341983317887,
+ "grad_norm": 139.85794067382812,
+ "learning_rate": 2.503097893432466e-05,
+ "loss": 8.6601,
+ "step": 670
+ },
+ {
+ "epoch": 2.5264133456904543,
+ "grad_norm": 160.73171997070312,
+ "learning_rate": 2.4907063197026023e-05,
+ "loss": 9.522,
+ "step": 680
+ },
+ {
+ "epoch": 2.5634847080630214,
+ "grad_norm": 190.7194366455078,
+ "learning_rate": 2.4783147459727385e-05,
+ "loss": 10.0619,
+ "step": 690
+ },
+ {
+ "epoch": 2.6005560704355886,
+ "grad_norm": 172.5655517578125,
+ "learning_rate": 2.4659231722428747e-05,
+ "loss": 9.6444,
+ "step": 700
+ },
+ {
+ "epoch": 2.6376274328081557,
+ "grad_norm": 280.9183044433594,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 10.7593,
+ "step": 710
+ },
+ {
+ "epoch": 2.674698795180723,
+ "grad_norm": 118.03450775146484,
+ "learning_rate": 2.4411400247831474e-05,
+ "loss": 10.4432,
+ "step": 720
+ },
+ {
+ "epoch": 2.71177015755329,
+ "grad_norm": 332.662353515625,
+ "learning_rate": 2.428748451053284e-05,
+ "loss": 9.6033,
+ "step": 730
+ },
+ {
+ "epoch": 2.748841519925857,
+ "grad_norm": 129.18788146972656,
+ "learning_rate": 2.41635687732342e-05,
+ "loss": 9.108,
+ "step": 740
+ },
+ {
+ "epoch": 2.7859128822984243,
+ "grad_norm": 350.5856628417969,
+ "learning_rate": 2.4039653035935564e-05,
+ "loss": 9.884,
+ "step": 750
+ },
+ {
+ "epoch": 2.7859128822984243,
+ "eval_loss": 2.599517583847046,
+ "eval_runtime": 34.0592,
+ "eval_samples_per_second": 47.506,
+ "eval_steps_per_second": 6.812,
+ "step": 750
+ },
+ {
+ "epoch": 2.822984244670992,
+ "grad_norm": 285.14862060546875,
+ "learning_rate": 2.391573729863693e-05,
+ "loss": 11.3952,
+ "step": 760
+ },
+ {
+ "epoch": 2.860055607043559,
+ "grad_norm": 185.5862579345703,
+ "learning_rate": 2.379182156133829e-05,
+ "loss": 10.6798,
+ "step": 770
+ },
+ {
+ "epoch": 2.897126969416126,
+ "grad_norm": 77.20765686035156,
+ "learning_rate": 2.3667905824039656e-05,
+ "loss": 9.9795,
+ "step": 780
+ },
+ {
+ "epoch": 2.9341983317886933,
+ "grad_norm": 88.75109100341797,
+ "learning_rate": 2.3543990086741015e-05,
+ "loss": 9.3694,
+ "step": 790
+ },
+ {
+ "epoch": 2.9712696941612604,
+ "grad_norm": 53.416690826416016,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 8.4631,
+ "step": 800
+ },
+ {
+ "epoch": 3.0111214087117704,
+ "grad_norm": 72.12966918945312,
+ "learning_rate": 2.3296158612143742e-05,
+ "loss": 9.3903,
+ "step": 810
+ },
+ {
+ "epoch": 3.0481927710843375,
+ "grad_norm": 159.3909912109375,
+ "learning_rate": 2.3172242874845104e-05,
+ "loss": 8.9608,
+ "step": 820
+ },
+ {
+ "epoch": 3.0852641334569046,
+ "grad_norm": 161.03713989257812,
+ "learning_rate": 2.304832713754647e-05,
+ "loss": 9.623,
+ "step": 830
+ },
+ {
+ "epoch": 3.122335495829472,
+ "grad_norm": 136.50045776367188,
+ "learning_rate": 2.2924411400247832e-05,
+ "loss": 8.5958,
+ "step": 840
+ },
+ {
+ "epoch": 3.159406858202039,
+ "grad_norm": 60.785926818847656,
+ "learning_rate": 2.2800495662949197e-05,
+ "loss": 8.6278,
+ "step": 850
+ },
+ {
+ "epoch": 3.196478220574606,
+ "grad_norm": 115.2509536743164,
+ "learning_rate": 2.267657992565056e-05,
+ "loss": 8.661,
+ "step": 860
+ },
+ {
+ "epoch": 3.233549582947173,
+ "grad_norm": 144.38912963867188,
+ "learning_rate": 2.255266418835192e-05,
+ "loss": 9.5836,
+ "step": 870
+ },
+ {
+ "epoch": 3.2706209453197403,
+ "grad_norm": 239.74520874023438,
+ "learning_rate": 2.2428748451053287e-05,
+ "loss": 8.6221,
+ "step": 880
+ },
+ {
+ "epoch": 3.3076923076923075,
+ "grad_norm": 215.12789916992188,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 9.1995,
+ "step": 890
+ },
+ {
+ "epoch": 3.344763670064875,
+ "grad_norm": 113.77183532714844,
+ "learning_rate": 2.218091697645601e-05,
+ "loss": 8.6909,
+ "step": 900
+ },
+ {
+ "epoch": 3.344763670064875,
+ "eval_loss": 2.4246325492858887,
+ "eval_runtime": 34.1351,
+ "eval_samples_per_second": 47.4,
+ "eval_steps_per_second": 6.797,
+ "step": 900
+ },
+ {
+ "epoch": 3.381835032437442,
+ "grad_norm": 175.02731323242188,
+ "learning_rate": 2.2057001239157373e-05,
+ "loss": 8.7632,
+ "step": 910
+ },
+ {
+ "epoch": 3.4189063948100094,
+ "grad_norm": 142.76580810546875,
+ "learning_rate": 2.1933085501858735e-05,
+ "loss": 8.5604,
+ "step": 920
+ },
+ {
+ "epoch": 3.4559777571825765,
+ "grad_norm": 163.0391387939453,
+ "learning_rate": 2.18091697645601e-05,
+ "loss": 9.3871,
+ "step": 930
+ },
+ {
+ "epoch": 3.4930491195551436,
+ "grad_norm": 386.4294738769531,
+ "learning_rate": 2.1685254027261462e-05,
+ "loss": 9.6445,
+ "step": 940
+ },
+ {
+ "epoch": 3.5301204819277108,
+ "grad_norm": 86.92581939697266,
+ "learning_rate": 2.1561338289962827e-05,
+ "loss": 9.635,
+ "step": 950
+ },
+ {
+ "epoch": 3.567191844300278,
+ "grad_norm": 80.94845581054688,
+ "learning_rate": 2.143742255266419e-05,
+ "loss": 8.8689,
+ "step": 960
+ },
+ {
+ "epoch": 3.6042632066728455,
+ "grad_norm": 69.99708557128906,
+ "learning_rate": 2.1313506815365555e-05,
+ "loss": 9.1411,
+ "step": 970
+ },
+ {
+ "epoch": 3.641334569045412,
+ "grad_norm": 80.53229522705078,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 8.7552,
+ "step": 980
+ },
+ {
+ "epoch": 3.67840593141798,
+ "grad_norm": 53.3475341796875,
+ "learning_rate": 2.1065675340768275e-05,
+ "loss": 8.5697,
+ "step": 990
+ },
+ {
+ "epoch": 3.715477293790547,
+ "grad_norm": 54.19875717163086,
+ "learning_rate": 2.094175960346964e-05,
+ "loss": 8.9008,
+ "step": 1000
+ },
+ {
+ "epoch": 3.752548656163114,
+ "grad_norm": 52.96628952026367,
+ "learning_rate": 2.0817843866171003e-05,
+ "loss": 8.9415,
+ "step": 1010
+ },
+ {
+ "epoch": 3.789620018535681,
+ "grad_norm": 64.08312225341797,
+ "learning_rate": 2.0693928128872368e-05,
+ "loss": 8.1303,
+ "step": 1020
+ },
+ {
+ "epoch": 3.8266913809082483,
+ "grad_norm": 68.22163391113281,
+ "learning_rate": 2.057001239157373e-05,
+ "loss": 8.241,
+ "step": 1030
+ },
+ {
+ "epoch": 3.8637627432808155,
+ "grad_norm": 66.70902252197266,
+ "learning_rate": 2.0446096654275092e-05,
+ "loss": 8.6969,
+ "step": 1040
+ },
+ {
+ "epoch": 3.9008341056533826,
+ "grad_norm": 49.94658279418945,
+ "learning_rate": 2.0322180916976458e-05,
+ "loss": 9.8392,
+ "step": 1050
+ },
+ {
+ "epoch": 3.9008341056533826,
+ "eval_loss": 2.533297300338745,
+ "eval_runtime": 34.1106,
+ "eval_samples_per_second": 47.434,
+ "eval_steps_per_second": 6.801,
+ "step": 1050
+ },
+ {
+ "epoch": 3.93790546802595,
+ "grad_norm": 73.9527816772461,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 8.5934,
+ "step": 1060
+ },
+ {
+ "epoch": 3.974976830398517,
+ "grad_norm": 76.90135955810547,
+ "learning_rate": 2.6703841387856257e-05,
+ "loss": 8.006,
+ "step": 1070
+ },
+ {
+ "epoch": 4.014828544949027,
+ "grad_norm": 85.86236572265625,
+ "learning_rate": 2.664188351920694e-05,
+ "loss": 9.0081,
+ "step": 1080
+ },
+ {
+ "epoch": 4.051899907321594,
+ "grad_norm": 106.57842254638672,
+ "learning_rate": 2.6579925650557622e-05,
+ "loss": 8.6706,
+ "step": 1090
+ },
+ {
+ "epoch": 4.088971269694161,
+ "grad_norm": 68.10344696044922,
+ "learning_rate": 2.6517967781908303e-05,
+ "loss": 9.6146,
+ "step": 1100
+ },
+ {
+ "epoch": 4.126042632066729,
+ "grad_norm": 67.02091217041016,
+ "learning_rate": 2.6456009913258984e-05,
+ "loss": 9.225,
+ "step": 1110
+ },
+ {
+ "epoch": 4.163113994439295,
+ "grad_norm": 67.66644287109375,
+ "learning_rate": 2.6394052044609665e-05,
+ "loss": 8.7522,
+ "step": 1120
+ },
+ {
+ "epoch": 4.200185356811863,
+ "grad_norm": 59.26814651489258,
+ "learning_rate": 2.633209417596035e-05,
+ "loss": 9.0221,
+ "step": 1130
+ },
+ {
+ "epoch": 4.23725671918443,
+ "grad_norm": 88.07310485839844,
+ "learning_rate": 2.627013630731103e-05,
+ "loss": 9.6458,
+ "step": 1140
+ },
+ {
+ "epoch": 4.274328081556997,
+ "grad_norm": 69.33895874023438,
+ "learning_rate": 2.620817843866171e-05,
+ "loss": 8.7692,
+ "step": 1150
+ },
+ {
+ "epoch": 4.311399443929565,
+ "grad_norm": 43.823951721191406,
+ "learning_rate": 2.6146220570012392e-05,
+ "loss": 9.2874,
+ "step": 1160
+ },
+ {
+ "epoch": 4.348470806302132,
+ "grad_norm": 63.590980529785156,
+ "learning_rate": 2.6084262701363077e-05,
+ "loss": 8.9276,
+ "step": 1170
+ },
+ {
+ "epoch": 4.385542168674699,
+ "grad_norm": 43.4519157409668,
+ "learning_rate": 2.6022304832713758e-05,
+ "loss": 8.7444,
+ "step": 1180
+ },
+ {
+ "epoch": 4.422613531047266,
+ "grad_norm": 72.1529312133789,
+ "learning_rate": 2.5960346964064435e-05,
+ "loss": 8.7265,
+ "step": 1190
+ },
+ {
+ "epoch": 4.459684893419833,
+ "grad_norm": 101.67008972167969,
+ "learning_rate": 2.5898389095415116e-05,
+ "loss": 8.7642,
+ "step": 1200
+ },
+ {
+ "epoch": 4.459684893419833,
+ "eval_loss": 2.6471314430236816,
+ "eval_runtime": 34.2723,
+ "eval_samples_per_second": 47.21,
+ "eval_steps_per_second": 6.769,
+ "step": 1200
+ },
+ {
+ "epoch": 4.4967562557924,
+ "grad_norm": 66.03606414794922,
+ "learning_rate": 2.5836431226765797e-05,
+ "loss": 8.8917,
+ "step": 1210
+ },
+ {
+ "epoch": 4.533827618164968,
+ "grad_norm": 60.01662826538086,
+ "learning_rate": 2.5774473358116482e-05,
+ "loss": 9.2155,
+ "step": 1220
+ },
+ {
+ "epoch": 4.570898980537534,
+ "grad_norm": 83.86510467529297,
+ "learning_rate": 2.5712515489467163e-05,
+ "loss": 8.6101,
+ "step": 1230
+ },
+ {
+ "epoch": 4.607970342910102,
+ "grad_norm": 64.35486602783203,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 8.9904,
+ "step": 1240
+ },
+ {
+ "epoch": 4.645041705282669,
+ "grad_norm": 60.44017791748047,
+ "learning_rate": 2.5588599752168525e-05,
+ "loss": 9.3272,
+ "step": 1250
+ },
+ {
+ "epoch": 4.682113067655236,
+ "grad_norm": 56.501102447509766,
+ "learning_rate": 2.5526641883519206e-05,
+ "loss": 7.9367,
+ "step": 1260
+ },
+ {
+ "epoch": 4.719184430027804,
+ "grad_norm": 65.01017761230469,
+ "learning_rate": 2.546468401486989e-05,
+ "loss": 8.5891,
+ "step": 1270
+ },
+ {
+ "epoch": 4.7562557924003706,
+ "grad_norm": 63.42623519897461,
+ "learning_rate": 2.540272614622057e-05,
+ "loss": 8.6286,
+ "step": 1280
+ },
+ {
+ "epoch": 4.793327154772938,
+ "grad_norm": 58.96747589111328,
+ "learning_rate": 2.5340768277571252e-05,
+ "loss": 7.9982,
+ "step": 1290
+ },
+ {
+ "epoch": 4.830398517145505,
+ "grad_norm": 96.77368927001953,
+ "learning_rate": 2.5278810408921933e-05,
+ "loss": 7.5587,
+ "step": 1300
+ },
+ {
+ "epoch": 4.867469879518072,
+ "grad_norm": 85.92269897460938,
+ "learning_rate": 2.5216852540272614e-05,
+ "loss": 7.9405,
+ "step": 1310
+ },
+ {
+ "epoch": 4.904541241890639,
+ "grad_norm": 86.37085723876953,
+ "learning_rate": 2.51548946716233e-05,
+ "loss": 9.7092,
+ "step": 1320
+ },
+ {
+ "epoch": 4.941612604263207,
+ "grad_norm": 65.88274383544922,
+ "learning_rate": 2.509293680297398e-05,
+ "loss": 8.1475,
+ "step": 1330
+ },
+ {
+ "epoch": 4.978683966635774,
+ "grad_norm": 53.252742767333984,
+ "learning_rate": 2.503097893432466e-05,
+ "loss": 9.3603,
+ "step": 1340
+ },
+ {
+ "epoch": 5.014828544949027,
+ "grad_norm": 182.35781860351562,
+ "learning_rate": 2.4969021065675342e-05,
+ "loss": 7.6621,
+ "step": 1350
+ },
+ {
+ "epoch": 5.014828544949027,
+ "eval_loss": 2.8308892250061035,
+ "eval_runtime": 34.1063,
+ "eval_samples_per_second": 47.44,
+ "eval_steps_per_second": 6.802,
+ "step": 1350
+ },
+ {
+ "epoch": 5.051899907321594,
+ "grad_norm": 324.6659240722656,
+ "learning_rate": 2.4907063197026023e-05,
+ "loss": 9.2301,
+ "step": 1360
+ },
+ {
+ "epoch": 5.088971269694161,
+ "grad_norm": 4839.4453125,
+ "learning_rate": 2.4845105328376707e-05,
+ "loss": 9.7789,
+ "step": 1370
+ },
+ {
+ "epoch": 5.126042632066729,
+ "grad_norm": 408588.40625,
+ "learning_rate": 2.4783147459727385e-05,
+ "loss": 9.5359,
+ "step": 1380
+ },
+ {
+ "epoch": 5.163113994439295,
+ "grad_norm": 932.5144653320312,
+ "learning_rate": 2.4721189591078066e-05,
+ "loss": 10.8065,
+ "step": 1390
+ },
+ {
+ "epoch": 5.200185356811863,
+ "grad_norm": 148.0634765625,
+ "learning_rate": 2.4659231722428747e-05,
+ "loss": 10.0149,
+ "step": 1400
+ },
+ {
+ "epoch": 5.23725671918443,
+ "grad_norm": 1214.7279052734375,
+ "learning_rate": 2.459727385377943e-05,
+ "loss": 10.2582,
+ "step": 1410
+ },
+ {
+ "epoch": 5.274328081556997,
+ "grad_norm": 188.14292907714844,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 10.16,
+ "step": 1420
+ },
+ {
+ "epoch": 5.311399443929565,
+ "grad_norm": 388.7134094238281,
+ "learning_rate": 2.4473358116480793e-05,
+ "loss": 10.0763,
+ "step": 1430
+ },
+ {
+ "epoch": 5.348470806302132,
+ "grad_norm": 270.61279296875,
+ "learning_rate": 2.4411400247831474e-05,
+ "loss": 9.5737,
+ "step": 1440
+ },
+ {
+ "epoch": 5.385542168674699,
+ "grad_norm": 1637.854736328125,
+ "learning_rate": 2.4349442379182155e-05,
+ "loss": 10.4816,
+ "step": 1450
+ },
+ {
+ "epoch": 5.422613531047266,
+ "grad_norm": 1118.7012939453125,
+ "learning_rate": 2.428748451053284e-05,
+ "loss": 8.6687,
+ "step": 1460
+ },
+ {
+ "epoch": 5.459684893419833,
+ "grad_norm": 788.9707641601562,
+ "learning_rate": 2.422552664188352e-05,
+ "loss": 8.4066,
+ "step": 1470
+ },
+ {
+ "epoch": 5.4967562557924,
+ "grad_norm": 185.22950744628906,
+ "learning_rate": 2.41635687732342e-05,
+ "loss": 9.386,
+ "step": 1480
+ },
+ {
+ "epoch": 5.533827618164968,
+ "grad_norm": 257.726806640625,
+ "learning_rate": 2.4101610904584883e-05,
+ "loss": 8.3911,
+ "step": 1490
+ },
+ {
+ "epoch": 5.570898980537534,
+ "grad_norm": 867.30908203125,
+ "learning_rate": 2.4039653035935564e-05,
+ "loss": 8.8025,
+ "step": 1500
+ },
+ {
+ "epoch": 5.570898980537534,
+ "eval_loss": 2.5407650470733643,
+ "eval_runtime": 34.0934,
+ "eval_samples_per_second": 47.458,
+ "eval_steps_per_second": 6.805,
+ "step": 1500
+ },
+ {
+ "epoch": 5.607970342910102,
+ "grad_norm": 68.35711669921875,
+ "learning_rate": 2.3977695167286248e-05,
+ "loss": 8.7939,
+ "step": 1510
+ },
+ {
+ "epoch": 5.645041705282669,
+ "grad_norm": 1419.1650390625,
+ "learning_rate": 2.391573729863693e-05,
+ "loss": 9.0903,
+ "step": 1520
+ },
+ {
+ "epoch": 5.682113067655236,
+ "grad_norm": 60.88732147216797,
+ "learning_rate": 2.385377942998761e-05,
+ "loss": 8.9878,
+ "step": 1530
+ },
+ {
+ "epoch": 5.719184430027804,
+ "grad_norm": 61.50148010253906,
+ "learning_rate": 2.379182156133829e-05,
+ "loss": 8.8642,
+ "step": 1540
+ },
+ {
+ "epoch": 5.7562557924003706,
+ "grad_norm": 73.9307861328125,
+ "learning_rate": 2.3729863692688972e-05,
+ "loss": 8.8625,
+ "step": 1550
+ },
+ {
+ "epoch": 5.793327154772938,
+ "grad_norm": 82.28329467773438,
+ "learning_rate": 2.3667905824039656e-05,
+ "loss": 8.4105,
+ "step": 1560
+ },
+ {
+ "epoch": 5.830398517145505,
+ "grad_norm": 83.54141235351562,
+ "learning_rate": 2.3605947955390337e-05,
+ "loss": 9.0163,
+ "step": 1570
+ },
+ {
+ "epoch": 5.867469879518072,
+ "grad_norm": 66.59397888183594,
+ "learning_rate": 2.3543990086741015e-05,
+ "loss": 8.8947,
+ "step": 1580
+ },
+ {
+ "epoch": 5.904541241890639,
+ "grad_norm": 55.8841667175293,
+ "learning_rate": 2.3482032218091696e-05,
+ "loss": 8.5647,
+ "step": 1590
+ },
+ {
+ "epoch": 5.941612604263207,
+ "grad_norm": 74.56892395019531,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 7.7047,
+ "step": 1600
+ },
+ {
+ "epoch": 5.978683966635774,
+ "grad_norm": 106.54405212402344,
+ "learning_rate": 2.335811648079306e-05,
+ "loss": 8.1484,
+ "step": 1610
+ },
+ {
+ "epoch": 6.014828544949027,
+ "grad_norm": 107.8216781616211,
+ "learning_rate": 2.3296158612143742e-05,
+ "loss": 8.4079,
+ "step": 1620
+ },
+ {
+ "epoch": 6.051899907321594,
+ "grad_norm": 100.98133087158203,
+ "learning_rate": 2.3234200743494423e-05,
+ "loss": 8.5027,
+ "step": 1630
+ },
+ {
+ "epoch": 6.088971269694161,
+ "grad_norm": 67.31327819824219,
+ "learning_rate": 2.3172242874845104e-05,
+ "loss": 8.1805,
+ "step": 1640
+ },
+ {
+ "epoch": 6.126042632066729,
+ "grad_norm": 61.10900115966797,
+ "learning_rate": 2.311028500619579e-05,
+ "loss": 8.4519,
+ "step": 1650
+ },
+ {
+ "epoch": 6.126042632066729,
+ "eval_loss": 2.590090036392212,
+ "eval_runtime": 34.1392,
+ "eval_samples_per_second": 47.394,
+ "eval_steps_per_second": 6.796,
+ "step": 1650
+ },
+ {
+ "epoch": 6.163113994439295,
+ "grad_norm": 60.91183090209961,
+ "learning_rate": 2.304832713754647e-05,
+ "loss": 9.062,
+ "step": 1660
+ },
+ {
+ "epoch": 6.200185356811863,
+ "grad_norm": 76.2873306274414,
+ "learning_rate": 2.298636926889715e-05,
+ "loss": 8.8499,
+ "step": 1670
+ },
+ {
+ "epoch": 6.23725671918443,
+ "grad_norm": 74.2361068725586,
+ "learning_rate": 2.2924411400247832e-05,
+ "loss": 8.6576,
+ "step": 1680
+ },
+ {
+ "epoch": 6.274328081556997,
+ "grad_norm": 84.99951934814453,
+ "learning_rate": 2.2862453531598513e-05,
+ "loss": 8.4652,
+ "step": 1690
+ },
+ {
+ "epoch": 6.311399443929565,
+ "grad_norm": 86.40947723388672,
+ "learning_rate": 2.2800495662949197e-05,
+ "loss": 9.0782,
+ "step": 1700
+ },
+ {
+ "epoch": 6.348470806302132,
+ "grad_norm": 75.45568084716797,
+ "learning_rate": 2.2738537794299878e-05,
+ "loss": 8.1532,
+ "step": 1710
+ },
+ {
+ "epoch": 6.385542168674699,
+ "grad_norm": 86.65908813476562,
+ "learning_rate": 2.267657992565056e-05,
+ "loss": 8.5185,
+ "step": 1720
+ },
+ {
+ "epoch": 6.422613531047266,
+ "grad_norm": 62.620758056640625,
+ "learning_rate": 2.261462205700124e-05,
+ "loss": 9.5908,
+ "step": 1730
+ },
+ {
+ "epoch": 6.459684893419833,
+ "grad_norm": 73.63095092773438,
+ "learning_rate": 2.255266418835192e-05,
+ "loss": 8.4188,
+ "step": 1740
+ },
+ {
+ "epoch": 6.4967562557924,
+ "grad_norm": 55.175045013427734,
+ "learning_rate": 2.2490706319702606e-05,
+ "loss": 8.1885,
+ "step": 1750
+ },
+ {
+ "epoch": 6.533827618164968,
+ "grad_norm": 47.647621154785156,
+ "learning_rate": 2.2428748451053287e-05,
+ "loss": 8.7666,
+ "step": 1760
+ },
+ {
+ "epoch": 6.570898980537534,
+ "grad_norm": 191.0419158935547,
+ "learning_rate": 2.2366790582403964e-05,
+ "loss": 8.6105,
+ "step": 1770
+ },
+ {
+ "epoch": 6.607970342910102,
+ "grad_norm": 57.81939697265625,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 8.664,
+ "step": 1780
+ },
+ {
+ "epoch": 6.645041705282669,
+ "grad_norm": 85.04927825927734,
+ "learning_rate": 2.2242874845105326e-05,
+ "loss": 8.5294,
+ "step": 1790
+ },
+ {
+ "epoch": 6.682113067655236,
+ "grad_norm": 55.604122161865234,
+ "learning_rate": 2.218091697645601e-05,
+ "loss": 9.1857,
+ "step": 1800
+ },
+ {
+ "epoch": 6.682113067655236,
+ "eval_loss": 2.497358560562134,
+ "eval_runtime": 34.0441,
+ "eval_samples_per_second": 47.527,
+ "eval_steps_per_second": 6.815,
+ "step": 1800
+ },
+ {
+ "epoch": 6.719184430027804,
+ "grad_norm": 75.21812438964844,
+ "learning_rate": 2.211895910780669e-05,
+ "loss": 8.7053,
+ "step": 1810
+ },
+ {
+ "epoch": 6.7562557924003706,
+ "grad_norm": 85.76676940917969,
+ "learning_rate": 2.2057001239157373e-05,
+ "loss": 8.1428,
+ "step": 1820
+ },
+ {
+ "epoch": 6.793327154772938,
+ "grad_norm": 69.85575866699219,
+ "learning_rate": 2.1995043370508054e-05,
+ "loss": 8.4988,
+ "step": 1830
+ },
+ {
+ "epoch": 6.830398517145505,
+ "grad_norm": 66.85325622558594,
+ "learning_rate": 2.1933085501858735e-05,
+ "loss": 8.4147,
+ "step": 1840
+ },
+ {
+ "epoch": 6.867469879518072,
+ "grad_norm": 60.645294189453125,
+ "learning_rate": 2.187112763320942e-05,
+ "loss": 9.069,
+ "step": 1850
+ },
+ {
+ "epoch": 6.904541241890639,
+ "grad_norm": 77.59921264648438,
+ "learning_rate": 2.18091697645601e-05,
+ "loss": 8.4405,
+ "step": 1860
+ },
+ {
+ "epoch": 6.941612604263207,
+ "grad_norm": 56.27996826171875,
+ "learning_rate": 2.174721189591078e-05,
+ "loss": 9.2157,
+ "step": 1870
+ },
+ {
+ "epoch": 6.978683966635774,
+ "grad_norm": 64.18400573730469,
+ "learning_rate": 2.1685254027261462e-05,
+ "loss": 9.5492,
+ "step": 1880
+ },
+ {
+ "epoch": 7.014828544949027,
+ "grad_norm": 76.78484344482422,
+ "learning_rate": 2.1623296158612143e-05,
+ "loss": 8.1325,
+ "step": 1890
+ },
+ {
+ "epoch": 7.051899907321594,
+ "grad_norm": 77.69871520996094,
+ "learning_rate": 2.1561338289962827e-05,
+ "loss": 8.324,
+ "step": 1900
+ },
+ {
+ "epoch": 7.088971269694161,
+ "grad_norm": 102.43340301513672,
+ "learning_rate": 2.149938042131351e-05,
+ "loss": 7.7097,
+ "step": 1910
+ },
+ {
+ "epoch": 7.126042632066729,
+ "grad_norm": 114.0189208984375,
+ "learning_rate": 2.143742255266419e-05,
+ "loss": 8.0982,
+ "step": 1920
+ },
+ {
+ "epoch": 7.163113994439295,
+ "grad_norm": 66.60838317871094,
+ "learning_rate": 2.137546468401487e-05,
+ "loss": 7.7669,
+ "step": 1930
+ },
+ {
+ "epoch": 7.200185356811863,
+ "grad_norm": 252.72503662109375,
+ "learning_rate": 2.1313506815365555e-05,
+ "loss": 7.809,
+ "step": 1940
+ },
+ {
+ "epoch": 7.23725671918443,
+ "grad_norm": 91.0379409790039,
+ "learning_rate": 2.1251548946716236e-05,
+ "loss": 7.9729,
+ "step": 1950
+ },
+ {
+ "epoch": 7.23725671918443,
+ "eval_loss": 2.610764741897583,
+ "eval_runtime": 34.0774,
+ "eval_samples_per_second": 47.48,
+ "eval_steps_per_second": 6.808,
+ "step": 1950
+ },
+ {
+ "epoch": 7.274328081556997,
+ "grad_norm": 96.09555053710938,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 8.2125,
+ "step": 1960
+ },
+ {
+ "epoch": 7.311399443929565,
+ "grad_norm": 68.51132202148438,
+ "learning_rate": 2.1127633209417594e-05,
+ "loss": 7.7403,
+ "step": 1970
+ },
+ {
+ "epoch": 7.348470806302132,
+ "grad_norm": 94.45308685302734,
+ "learning_rate": 2.1065675340768275e-05,
+ "loss": 7.5494,
+ "step": 1980
+ },
+ {
+ "epoch": 7.385542168674699,
+ "grad_norm": 56.785587310791016,
+ "learning_rate": 2.100371747211896e-05,
+ "loss": 8.2821,
+ "step": 1990
+ },
+ {
+ "epoch": 7.422613531047266,
+ "grad_norm": 91.59173583984375,
+ "learning_rate": 2.094175960346964e-05,
+ "loss": 8.1644,
+ "step": 2000
+ },
+ {
+ "epoch": 7.459684893419833,
+ "grad_norm": 78.57572174072266,
+ "learning_rate": 2.0879801734820322e-05,
+ "loss": 8.1664,
+ "step": 2010
+ },
+ {
+ "epoch": 7.4967562557924,
+ "grad_norm": 71.83062744140625,
+ "learning_rate": 2.0817843866171003e-05,
+ "loss": 8.5876,
+ "step": 2020
+ },
+ {
+ "epoch": 7.533827618164968,
+ "grad_norm": 74.75129699707031,
+ "learning_rate": 2.0755885997521684e-05,
+ "loss": 8.2753,
+ "step": 2030
+ },
+ {
+ "epoch": 7.570898980537534,
+ "grad_norm": 96.26950073242188,
+ "learning_rate": 2.0693928128872368e-05,
+ "loss": 9.2057,
+ "step": 2040
+ },
+ {
+ "epoch": 7.607970342910102,
+ "grad_norm": 96.87144470214844,
+ "learning_rate": 2.063197026022305e-05,
+ "loss": 8.0052,
+ "step": 2050
+ },
+ {
+ "epoch": 7.645041705282669,
+ "grad_norm": 83.44422912597656,
+ "learning_rate": 2.057001239157373e-05,
+ "loss": 8.4954,
+ "step": 2060
+ },
+ {
+ "epoch": 7.682113067655236,
+ "grad_norm": 86.1534194946289,
+ "learning_rate": 2.050805452292441e-05,
+ "loss": 8.0325,
+ "step": 2070
+ },
+ {
+ "epoch": 7.719184430027804,
+ "grad_norm": 89.11744689941406,
+ "learning_rate": 2.0446096654275092e-05,
+ "loss": 8.2934,
+ "step": 2080
+ },
+ {
+ "epoch": 7.7562557924003706,
+ "grad_norm": 60.81666946411133,
+ "learning_rate": 2.0384138785625777e-05,
+ "loss": 9.4019,
+ "step": 2090
+ },
+ {
+ "epoch": 7.793327154772938,
+ "grad_norm": 82.93228912353516,
+ "learning_rate": 2.0322180916976458e-05,
+ "loss": 8.874,
+ "step": 2100
+ },
+ {
+ "epoch": 7.793327154772938,
+ "eval_loss": 2.452894926071167,
+ "eval_runtime": 34.0684,
+ "eval_samples_per_second": 47.493,
+ "eval_steps_per_second": 6.81,
+ "step": 2100
+ },
+ {
+ "epoch": 7.830398517145505,
+ "grad_norm": 62.004817962646484,
+ "learning_rate": 2.026022304832714e-05,
+ "loss": 8.0985,
+ "step": 2110
+ },
+ {
+ "epoch": 7.867469879518072,
+ "grad_norm": 72.79752349853516,
+ "learning_rate": 2.019826517967782e-05,
+ "loss": 8.9063,
+ "step": 2120
+ },
+ {
+ "epoch": 7.904541241890639,
+ "grad_norm": 1507.2855224609375,
+ "learning_rate": 2.01363073110285e-05,
+ "loss": 8.6332,
+ "step": 2130
+ },
+ {
+ "epoch": 7.941612604263207,
+ "grad_norm": 134.11378479003906,
+ "learning_rate": 2.0074349442379185e-05,
+ "loss": 9.4106,
+ "step": 2140
+ },
+ {
+ "epoch": 7.978683966635774,
+ "grad_norm": 72.19046783447266,
+ "learning_rate": 2.0012391573729866e-05,
+ "loss": 8.2304,
+ "step": 2150
+ },
+ {
+ "epoch": 8.014828544949028,
+ "grad_norm": 57.130165100097656,
+ "learning_rate": 1.9950433705080544e-05,
+ "loss": 7.9675,
+ "step": 2160
+ },
+ {
+ "epoch": 8.051899907321594,
+ "grad_norm": 72.58036041259766,
+ "learning_rate": 1.9888475836431225e-05,
+ "loss": 7.602,
+ "step": 2170
+ },
+ {
+ "epoch": 8.088971269694161,
+ "grad_norm": 91.77513885498047,
+ "learning_rate": 1.982651796778191e-05,
+ "loss": 7.9151,
+ "step": 2180
+ },
+ {
+ "epoch": 8.126042632066728,
+ "grad_norm": 111.52696228027344,
+ "learning_rate": 1.976456009913259e-05,
+ "loss": 7.5653,
+ "step": 2190
+ },
+ {
+ "epoch": 8.163113994439296,
+ "grad_norm": 74.9671401977539,
+ "learning_rate": 1.970260223048327e-05,
+ "loss": 7.1567,
+ "step": 2200
+ },
+ {
+ "epoch": 8.200185356811863,
+ "grad_norm": 124.09498596191406,
+ "learning_rate": 1.9640644361833952e-05,
+ "loss": 8.1708,
+ "step": 2210
+ },
+ {
+ "epoch": 8.23725671918443,
+ "grad_norm": 95.21754455566406,
+ "learning_rate": 1.9578686493184633e-05,
+ "loss": 7.0013,
+ "step": 2220
+ },
+ {
+ "epoch": 8.274328081556996,
+ "grad_norm": 135.3839569091797,
+ "learning_rate": 1.9516728624535318e-05,
+ "loss": 7.5999,
+ "step": 2230
+ },
+ {
+ "epoch": 8.311399443929565,
+ "grad_norm": 89.70751190185547,
+ "learning_rate": 1.9454770755886e-05,
+ "loss": 7.2028,
+ "step": 2240
+ },
+ {
+ "epoch": 8.348470806302132,
+ "grad_norm": 70.08137512207031,
+ "learning_rate": 1.939281288723668e-05,
+ "loss": 8.3759,
+ "step": 2250
+ },
+ {
+ "epoch": 8.348470806302132,
+ "eval_loss": 2.746619939804077,
+ "eval_runtime": 34.0831,
+ "eval_samples_per_second": 47.472,
+ "eval_steps_per_second": 6.807,
+ "step": 2250
+ },
+ {
+ "epoch": 8.400370713623726,
+ "grad_norm": 69.07305145263672,
+ "learning_rate": 1.933085501858736e-05,
+ "loss": 8.9083,
+ "step": 2260
+ },
+ {
+ "epoch": 8.437442075996293,
+ "grad_norm": 66.43407440185547,
+ "learning_rate": 1.926889714993804e-05,
+ "loss": 6.9349,
+ "step": 2270
+ },
+ {
+ "epoch": 8.47451343836886,
+ "grad_norm": 167.85784912109375,
+ "learning_rate": 1.9206939281288726e-05,
+ "loss": 7.5041,
+ "step": 2280
+ },
+ {
+ "epoch": 8.511584800741428,
+ "grad_norm": 173.38304138183594,
+ "learning_rate": 1.9144981412639407e-05,
+ "loss": 7.3744,
+ "step": 2290
+ },
+ {
+ "epoch": 8.548656163113995,
+ "grad_norm": 80.71511840820312,
+ "learning_rate": 1.9083023543990088e-05,
+ "loss": 8.6541,
+ "step": 2300
+ },
+ {
+ "epoch": 8.585727525486561,
+ "grad_norm": 82.77088928222656,
+ "learning_rate": 1.902106567534077e-05,
+ "loss": 8.6305,
+ "step": 2310
+ },
+ {
+ "epoch": 8.62279888785913,
+ "grad_norm": 55.3135871887207,
+ "learning_rate": 1.895910780669145e-05,
+ "loss": 8.2577,
+ "step": 2320
+ },
+ {
+ "epoch": 8.659870250231696,
+ "grad_norm": 64.61204528808594,
+ "learning_rate": 1.8897149938042134e-05,
+ "loss": 7.6382,
+ "step": 2330
+ },
+ {
+ "epoch": 8.696941612604263,
+ "grad_norm": 77.9918212890625,
+ "learning_rate": 1.8835192069392815e-05,
+ "loss": 8.114,
+ "step": 2340
+ },
+ {
+ "epoch": 8.73401297497683,
+ "grad_norm": 50.139678955078125,
+ "learning_rate": 1.8773234200743496e-05,
+ "loss": 7.8875,
+ "step": 2350
+ },
+ {
+ "epoch": 8.771084337349398,
+ "grad_norm": 241.43588256835938,
+ "learning_rate": 1.8711276332094174e-05,
+ "loss": 7.0444,
+ "step": 2360
+ },
+ {
+ "epoch": 8.808155699721965,
+ "grad_norm": 133.25645446777344,
+ "learning_rate": 1.8649318463444855e-05,
+ "loss": 7.7393,
+ "step": 2370
+ },
+ {
+ "epoch": 8.845227062094532,
+ "grad_norm": 94.1192855834961,
+ "learning_rate": 1.858736059479554e-05,
+ "loss": 8.8284,
+ "step": 2380
+ },
+ {
+ "epoch": 8.882298424467098,
+ "grad_norm": 83.68805694580078,
+ "learning_rate": 1.852540272614622e-05,
+ "loss": 7.997,
+ "step": 2390
+ },
+ {
+ "epoch": 8.919369786839667,
+ "grad_norm": 67.01426696777344,
+ "learning_rate": 1.84634448574969e-05,
+ "loss": 7.786,
+ "step": 2400
+ },
+ {
+ "epoch": 8.919369786839667,
+ "eval_loss": 2.6791422367095947,
+ "eval_runtime": 34.2417,
+ "eval_samples_per_second": 47.252,
+ "eval_steps_per_second": 6.775,
+ "step": 2400
+ },
+ {
+ "epoch": 8.956441149212234,
+ "grad_norm": 72.36073303222656,
+ "learning_rate": 1.8401486988847582e-05,
+ "loss": 7.6257,
+ "step": 2410
+ },
+ {
+ "epoch": 8.9935125115848,
+ "grad_norm": 70.3509521484375,
+ "learning_rate": 1.8339529120198267e-05,
+ "loss": 7.099,
+ "step": 2420
+ },
+ {
+ "epoch": 9.03336422613531,
+ "grad_norm": 136.34727478027344,
+ "learning_rate": 1.8277571251548948e-05,
+ "loss": 8.041,
+ "step": 2430
+ },
+ {
+ "epoch": 9.070435588507877,
+ "grad_norm": 81.48332214355469,
+ "learning_rate": 1.821561338289963e-05,
+ "loss": 7.8606,
+ "step": 2440
+ },
+ {
+ "epoch": 9.107506950880445,
+ "grad_norm": 93.1724853515625,
+ "learning_rate": 1.815365551425031e-05,
+ "loss": 7.8551,
+ "step": 2450
+ },
+ {
+ "epoch": 9.144578313253012,
+ "grad_norm": 132.82778930664062,
+ "learning_rate": 1.809169764560099e-05,
+ "loss": 7.3977,
+ "step": 2460
+ },
+ {
+ "epoch": 9.181649675625579,
+ "grad_norm": 102.73971557617188,
+ "learning_rate": 1.8029739776951675e-05,
+ "loss": 7.8721,
+ "step": 2470
+ },
+ {
+ "epoch": 9.218721037998147,
+ "grad_norm": 106.7470932006836,
+ "learning_rate": 1.7967781908302356e-05,
+ "loss": 7.5839,
+ "step": 2480
+ },
+ {
+ "epoch": 9.255792400370714,
+ "grad_norm": 84.64981842041016,
+ "learning_rate": 1.7905824039653037e-05,
+ "loss": 7.1823,
+ "step": 2490
+ },
+ {
+ "epoch": 9.29286376274328,
+ "grad_norm": 85.59486389160156,
+ "learning_rate": 1.7843866171003718e-05,
+ "loss": 7.5513,
+ "step": 2500
+ },
+ {
+ "epoch": 9.329935125115847,
+ "grad_norm": 89.53803253173828,
+ "learning_rate": 1.77819083023544e-05,
+ "loss": 8.0879,
+ "step": 2510
+ },
+ {
+ "epoch": 9.367006487488416,
+ "grad_norm": 43.886898040771484,
+ "learning_rate": 1.7719950433705084e-05,
+ "loss": 7.5694,
+ "step": 2520
+ },
+ {
+ "epoch": 9.404077849860982,
+ "grad_norm": 73.4898910522461,
+ "learning_rate": 1.7657992565055765e-05,
+ "loss": 7.3436,
+ "step": 2530
+ },
+ {
+ "epoch": 9.44114921223355,
+ "grad_norm": 100.42314147949219,
+ "learning_rate": 1.7596034696406446e-05,
+ "loss": 6.9425,
+ "step": 2540
+ },
+ {
+ "epoch": 9.478220574606116,
+ "grad_norm": 121.53675079345703,
+ "learning_rate": 1.7534076827757123e-05,
+ "loss": 7.9461,
+ "step": 2550
+ },
+ {
+ "epoch": 9.478220574606116,
+ "eval_loss": 2.6608545780181885,
+ "eval_runtime": 34.2437,
+ "eval_samples_per_second": 47.25,
+ "eval_steps_per_second": 6.775,
+ "step": 2550
+ }
+ ],
+ "logging_steps": 10,
+ "max_steps": 5380,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 20,
+ "save_steps": 150,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 0.0,
+ "train_batch_size": 7,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-2550/training_args.bin b/checkpoints/checkpoint-2550/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..f1a69c7344d84f80932397ffdc7afef25687d5b4
--- /dev/null
+++ b/checkpoints/checkpoint-2550/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fda94467724e48ac975735648aa63a75558d48b6c993506e918a3476acf80697
+size 129
diff --git a/checkpoints/checkpoint-3000/1_Pooling/config.json b/checkpoints/checkpoint-3000/1_Pooling/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b68441c3f37d8bee79501ca0afe536bf19753928
--- /dev/null
+++ b/checkpoints/checkpoint-3000/1_Pooling/config.json
@@ -0,0 +1,10 @@
+{
+ "word_embedding_dimension": 1024,
+ "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
+}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-3000/README.md b/checkpoints/checkpoint-3000/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..fd1fe437cf7dce0dd2c6dc2aceaf3615448cfb24
--- /dev/null
+++ b/checkpoints/checkpoint-3000/README.md
@@ -0,0 +1,854 @@
+---
+tags:
+- sentence-transformers
+- sentence-similarity
+- feature-extraction
+- generated_from_trainer
+- dataset_size:7552
+- loss:CoSENTLoss
+base_model: intfloat/multilingual-e5-large-instruct
+widget:
+- source_sentence: How are calibration points linked to equipment?
+ sentences:
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+ - "How does a flow computer generate and store reports?\nA flow computer generates\
+ \ daily or hourly reports to provide users with operational data. These reports\
+ \ are stored in the flow computer's memory in an organized format.\n\nReport structure:\n\
+ - Each report includes:\n- Date and time of the data recording.\n- Data recorded\
+ \ from flow computers.\n\nData storage in tables:\nThe reports are saved in two\
+ \ tables:\n1. Main table (Index):\n - Stores the date, time, and flow computer\
+ \ identifier.\n2. Detail table:\n - Stores the measured values associated with\
+ \ the report.\n\nConnection to the Modbus table:\nThe flow computer's reports\
+ \ are linked to a Modbus table. This table contains the names corresponding to\
+ \ each value in the reports, making it easier to interpret the data."
+ - "What is uncertainty?\nUncertainty is a measure of confidence in the precision\
+ \ and reliability of results obtained from equipment or measurement systems. It\
+ \ quantifies the potential error or margin of error in measurements.\n\nTypes\
+ \ of uncertainty:\nThere are two main types of uncertainty:\n1. Uncertainty of\
+ \ magnitudes (variables):\n - Refers to the uncertainty of specific variables,\
+ \ such as temperature or pressure.\n - It is calculated after calibrating a\
+ \ device or obtained from the equipment manufacturer's manual.\n - This uncertainty\
+ \ serves as a starting point for further calculations related to the equipment.\n\
+ \n2. Uncertainty of the measurement system:\n - Refers to the uncertainty calculated\
+ \ for the overall flow measurement.\n - It depends on the uncertainties of\
+ \ the individual variables (magnitudes) and represents the combined margin of\
+ \ error for the entire system.\n\nKey points:\n- The uncertainties of magnitudes\
+ \ (variables) are the foundation for calculating the uncertainty of the measurement\
+ \ system. Think of them as the \"building blocks.\"\n- Do not confuse the two\
+ \ types of uncertainty:\n - **Uncertainty of magnitudes/variables**: Specific\
+ \ to individual variables (e.g., temperature, pressure).\n - **Uncertainty\
+ \ of the measurement system**: Specific to the overall flow measurement.\n\nDatabase\
+ \ storage for uncertainties:\nIn the database, uncertainty calculations are stored\
+ \ in two separate tables:\n1. Uncertainty of magnitudes (variables):\n - Stores\
+ \ the uncertainty values for specific variables (e.g., temperature, pressure).\n\
+ \n2. Uncertainty of the measurement system:\n - Stores the uncertainty values\
+ \ for the overall flow measurement system.\n\nHow to retrieve uncertainty data:\n\
+ - To find the uncertainty of the measurement system, join the measurement systems\
+ \ table with the uncertainty of the measurement system table.\n- To find the uncertainty\
+ \ of a specific variable (magnitude), join the measurement systems table with\
+ \ the uncertainty of magnitudes (variables) table.\n\nImportant note:\nDo not\
+ \ confuse the two types of uncertainty:\n- If the user requests the uncertainty\
+ \ of the measurement system, use the first join (measurement systems table + uncertainty\
+ \ of the measurement system table).\n- If the user requests the uncertainty of\
+ \ a specific variable (magnitude) in a report, use the second join (measurement\
+ \ systems table + uncertainty of magnitudes table)."
+- source_sentence: What is the primary key of the flow computer table?
+ sentences:
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - "How does a flow computer generate and store reports?\nA flow computer generates\
+ \ daily or hourly reports to provide users with operational data. These reports\
+ \ are stored in the flow computer's memory in an organized format.\n\nReport structure:\n\
+ - Each report includes:\n- Date and time of the data recording.\n- Data recorded\
+ \ from flow computers.\n\nData storage in tables:\nThe reports are saved in two\
+ \ tables:\n1. Main table (Index):\n - Stores the date, time, and flow computer\
+ \ identifier.\n2. Detail table:\n - Stores the measured values associated with\
+ \ the report.\n\nConnection to the Modbus table:\nThe flow computer's reports\
+ \ are linked to a Modbus table. This table contains the names corresponding to\
+ \ each value in the reports, making it easier to interpret the data."
+- source_sentence: Can you provide a sample query to test the retrieval of the uncertainty
+ result for the specified tag and date?
+ sentences:
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - "What is uncertainty?\nUncertainty is a measure of confidence in the precision\
+ \ and reliability of results obtained from equipment or measurement systems. It\
+ \ quantifies the potential error or margin of error in measurements.\n\nTypes\
+ \ of uncertainty:\nThere are two main types of uncertainty:\n1. Uncertainty of\
+ \ magnitudes (variables):\n - Refers to the uncertainty of specific variables,\
+ \ such as temperature or pressure.\n - It is calculated after calibrating a\
+ \ device or obtained from the equipment manufacturer's manual.\n - This uncertainty\
+ \ serves as a starting point for further calculations related to the equipment.\n\
+ \n2. Uncertainty of the measurement system:\n - Refers to the uncertainty calculated\
+ \ for the overall flow measurement.\n - It depends on the uncertainties of\
+ \ the individual variables (magnitudes) and represents the combined margin of\
+ \ error for the entire system.\n\nKey points:\n- The uncertainties of magnitudes\
+ \ (variables) are the foundation for calculating the uncertainty of the measurement\
+ \ system. Think of them as the \"building blocks.\"\n- Do not confuse the two\
+ \ types of uncertainty:\n - **Uncertainty of magnitudes/variables**: Specific\
+ \ to individual variables (e.g., temperature, pressure).\n - **Uncertainty\
+ \ of the measurement system**: Specific to the overall flow measurement.\n\nDatabase\
+ \ storage for uncertainties:\nIn the database, uncertainty calculations are stored\
+ \ in two separate tables:\n1. Uncertainty of magnitudes (variables):\n - Stores\
+ \ the uncertainty values for specific variables (e.g., temperature, pressure).\n\
+ \n2. Uncertainty of the measurement system:\n - Stores the uncertainty values\
+ \ for the overall flow measurement system.\n\nHow to retrieve uncertainty data:\n\
+ - To find the uncertainty of the measurement system, join the measurement systems\
+ \ table with the uncertainty of the measurement system table.\n- To find the uncertainty\
+ \ of a specific variable (magnitude), join the measurement systems table with\
+ \ the uncertainty of magnitudes (variables) table.\n\nImportant note:\nDo not\
+ \ confuse the two types of uncertainty:\n- If the user requests the uncertainty\
+ \ of the measurement system, use the first join (measurement systems table + uncertainty\
+ \ of the measurement system table).\n- If the user requests the uncertainty of\
+ \ a specific variable (magnitude) in a report, use the second join (measurement\
+ \ systems table + uncertainty of magnitudes table)."
+- source_sentence: How are the secondary equipment and measurement system related?
+ sentences:
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - 'What do measurement equipment measure?
+
+ Each equipment measures a physical magnitude, also known as a variable. Based
+ on the type of variable they measure, devices are classified into different categories.
+
+
+ Equipment classification:
+
+ - Primary meter: Assigned by default to equipments like orifice plates.
+
+ - Secondary meter: Assigned by default to equipments like transmitters.
+
+ - Tertiary meter: Used for other types of equipments.
+
+
+ Equipment types in the database:
+
+ The database includes a table listing all equipment types. Examples of equipment
+ types are:
+
+ - Differential pressure transmitters
+
+ - RTDs (Resistance Temperature Detectors)
+
+ - Orifice plates
+
+ - Multivariable transmitters
+
+ - Ultrasonic meters
+
+
+ Meteorological checks for equipments:
+
+ Each equipment type is assigned a meteorological check, which can be either:
+
+ - Calibration: To ensure measurement accuracy.
+
+ - Inspection: To verify proper functioning.
+
+
+ Data storage in tables:
+
+ The database also includes a separate table for equipment classifications, which
+ are:
+
+ - Primary meter
+
+ - Secondary meter
+
+ - Tertiary meter
+
+ So, an equipment has equipment types and this types has classifications.'
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+- source_sentence: What is the table structure for secondary equipment?
+ sentences:
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+datasets:
+- Lauther/measuring-embeddings-v3
+pipeline_tag: sentence-similarity
+library_name: sentence-transformers
+---
+
+# SentenceTransformer based on intfloat/multilingual-e5-large-instruct
+
+This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [intfloat/multilingual-e5-large-instruct](https://huggingface.co/intfloat/multilingual-e5-large-instruct) on the [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) 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.
+
+## Model Details
+
+### Model Description
+- **Model Type:** Sentence Transformer
+- **Base model:** [intfloat/multilingual-e5-large-instruct](https://huggingface.co/intfloat/multilingual-e5-large-instruct)
+- **Maximum Sequence Length:** 512 tokens
+- **Output Dimensionality:** 1024 dimensions
+- **Similarity Function:** Cosine Similarity
+- **Training Dataset:**
+ - [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3)
+
+
+
+### Model Sources
+
+- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
+- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
+- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
+
+### Full Model Architecture
+
+```
+SentenceTransformer(
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: XLMRobertaModel
+ (1): Pooling({'word_embedding_dimension': 1024, '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()
+)
+```
+
+## Usage
+
+### Direct Usage (Sentence Transformers)
+
+First install the Sentence Transformers library:
+
+```bash
+pip install -U sentence-transformers
+```
+
+Then you can load this model and run inference.
+```python
+from sentence_transformers import SentenceTransformer
+
+# Download from the 🤗 Hub
+model = SentenceTransformer("sentence_transformers_model_id")
+# Run inference
+sentences = [
+ 'What is the table structure for secondary equipment?',
+ 'How are flow computers and measurement systems related?\nFlow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.\n\nDatabase terminology:\nIn the database, this relationship is referred to as:\n- Meter streams\n- Meter runs\n- Sections\n\nStorage of the relationship:\nThe relationship between a flow computer and its assigned measurement system is stored in a special table.\n\nUser context:\nWhen a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer.',
+ 'What kind of data store an equipment?\nEquipments can capture meteorological data, such as pressure, temperature, and volume (magnitudes). This data is essential for users to perform various calculations.\n\nData storage:\n- The measured values are stored in a special table in the database for magnitudes. This table contains the values of the variables captured by the equipments.\n- These values are **direct measurements** from the fluid (e.g., raw pressure, temperature, or volume readings). **They are not calculated values**, such as uncertainty.\n- The values stored in the variable values table are **different** from variable uncertainty values, which are calculated separately and represent the margin of error.\n\nAccessing the data:\n- Users typically access the data by referring to the readings from the measurement system, not directly from the individual equipments.\n- The readings are stored in a "variable values" table within the database.\n\nLinking variable names:\nIf the user needs to know the name of a variable, they must link the data to another table that stores information about the types of variables.',
+]
+embeddings = model.encode(sentences)
+print(embeddings.shape)
+# [3, 1024]
+
+# Get the similarity scores for the embeddings
+similarities = model.similarity(embeddings, embeddings)
+print(similarities.shape)
+# [3, 3]
+```
+
+
+
+
+
+
+
+
+
+
+
+## Training Details
+
+### Training Dataset
+
+#### measuring-embeddings-v3
+
+* Dataset: [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) at [1b3cbbe](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3/tree/1b3cbbeb70b63338110491cd3de2950fb40b4f87)
+* Size: 7,552 training samples
+* Columns: sentence1, sentence2, and score
+* Approximate statistics based on the first 1000 samples:
+ | | sentence1 | sentence2 | score |
+ |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:----------------------------------------------------------------|
+ | type | string | string | float |
+ | details | - min: 9 tokens
- mean: 15.96 tokens
- max: 40 tokens
| - min: 120 tokens
- mean: 255.56 tokens
- max: 512 tokens
| - min: 0.0
- mean: 0.22
- max: 0.95
|
+* Samples:
+ | sentence1 | sentence2 | score |
+ |:-------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------|
+ | How can I combine the sub-query with the main query to fetch the last uncertainty report? | What do measurement equipment measure?
Each equipment measures a physical magnitude, also known as a variable. Based on the type of variable they measure, devices are classified into different categories.
Equipment classification:
- Primary meter: Assigned by default to equipments like orifice plates.
- Secondary meter: Assigned by default to equipments like transmitters.
- Tertiary meter: Used for other types of equipments.
Equipment types in the database:
The database includes a table listing all equipment types. Examples of equipment types are:
- Differential pressure transmitters
- RTDs (Resistance Temperature Detectors)
- Orifice plates
- Multivariable transmitters
- Ultrasonic meters
Meteorological checks for equipments:
Each equipment type is assigned a meteorological check, which can be either:
- Calibration: To ensure measurement accuracy.
- Inspection: To verify proper functioning.
Data storage in tables:
The database also includes a separate table for equipment classific... | 0.1 |
+ | What is the column name for the calibration date in the calibration table? | How are flow computers and measurement systems related?
Flow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.
Database terminology:
In the database, this relationship is referred to as:
- Meter streams
- Meter runs
- Sections
Storage of the relationship:
The relationship between a flow computer and its assigned measurement system is stored in a special table.
User context:
When a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer. | 0.1 |
+ | What is the name of the table that contains the flow computer tags? | What is equipment calibration?
Calibration is a metrological verification process used to ensure the accuracy of measurement equipment. It is performed periodically, based on intervals set by the company or a regulatory body.
Purpose of calibration:
The calibration process corrects any deviations in how the equipment measures physical magnitudes (variables). This ensures the equipment provides accurate and reliable data.
Calibration cycles:
There are two main calibration cycles:
1. As-found: Represents the equipment's measurement accuracy before any adjustments are made. This cycle is almost always implemented.
2. As-left: Represents the equipment's measurement accuracy after adjustments are made. This cycle is used depending on regulatory requirements.
Calibration uncertainty:
- Uncertainty is included in the results of a calibration.
- Calibration uncertainty refers to the margin of error in the device's measurements, which also affects the uncertainty of the measured variable or ... | 0.05 |
+* Loss: [CoSENTLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
+ ```json
+ {
+ "scale": 20.0,
+ "similarity_fct": "pairwise_cos_sim"
+ }
+ ```
+
+### Evaluation Dataset
+
+#### measuring-embeddings-v3
+
+* Dataset: [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) at [1b3cbbe](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3/tree/1b3cbbeb70b63338110491cd3de2950fb40b4f87)
+* Size: 1,618 evaluation samples
+* Columns: sentence1, sentence2, and score
+* Approximate statistics based on the first 1000 samples:
+ | | sentence1 | sentence2 | score |
+ |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:----------------------------------------------------------------|
+ | type | string | string | float |
+ | details | - min: 9 tokens
- mean: 15.83 tokens
- max: 40 tokens
| - min: 120 tokens
- mean: 250.41 tokens
- max: 512 tokens
| - min: 0.0
- mean: 0.23
- max: 0.95
|
+* Samples:
+ | sentence1 | sentence2 | score |
+ |:--------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------|
+ | Identify any additional tables or columns that might be needed for the query. | How are flow computers and measurement systems related?
Flow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.
Database terminology:
In the database, this relationship is referred to as:
- Meter streams
- Meter runs
- Sections
Storage of the relationship:
The relationship between a flow computer and its assigned measurement system is stored in a special table.
User context:
When a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer. | 0.2 |
+ | What columns in these tables contain the measurement system tag and the flow computer tag? | How does a flow computer generate and store reports?
A flow computer generates daily or hourly reports to provide users with operational data. These reports are stored in the flow computer's memory in an organized format.
Report structure:
- Each report includes:
- Date and time of the data recording.
- Data recorded from flow computers.
Data storage in tables:
The reports are saved in two tables:
1. Main table (Index):
- Stores the date, time, and flow computer identifier.
2. Detail table:
- Stores the measured values associated with the report.
Connection to the Modbus table:
The flow computer's reports are linked to a Modbus table. This table contains the names corresponding to each value in the reports, making it easier to interpret the data. | 0.1 |
+ | Identify the column that stores the calibration number. | What kind of data store an equipment?
Equipments can capture meteorological data, such as pressure, temperature, and volume (magnitudes). This data is essential for users to perform various calculations.
Data storage:
- The measured values are stored in a special table in the database for magnitudes. This table contains the values of the variables captured by the equipments.
- These values are **direct measurements** from the fluid (e.g., raw pressure, temperature, or volume readings). **They are not calculated values**, such as uncertainty.
- The values stored in the variable values table are **different** from variable uncertainty values, which are calculated separately and represent the margin of error.
Accessing the data:
- Users typically access the data by referring to the readings from the measurement system, not directly from the individual equipments.
- The readings are stored in a "variable values" table within the database.
Linking variable names:
If the user needs to kno... | 0.1 |
+* Loss: [CoSENTLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
+ ```json
+ {
+ "scale": 20.0,
+ "similarity_fct": "pairwise_cos_sim"
+ }
+ ```
+
+### Training Hyperparameters
+#### Non-Default Hyperparameters
+
+- `eval_strategy`: steps
+- `per_device_train_batch_size`: 7
+- `per_device_eval_batch_size`: 7
+- `gradient_accumulation_steps`: 4
+- `learning_rate`: 3e-05
+- `num_train_epochs`: 20
+- `warmup_ratio`: 0.1
+
+#### All Hyperparameters
+Click to expand
+
+- `overwrite_output_dir`: False
+- `do_predict`: False
+- `eval_strategy`: steps
+- `prediction_loss_only`: True
+- `per_device_train_batch_size`: 7
+- `per_device_eval_batch_size`: 7
+- `per_gpu_train_batch_size`: None
+- `per_gpu_eval_batch_size`: None
+- `gradient_accumulation_steps`: 4
+- `eval_accumulation_steps`: None
+- `torch_empty_cache_steps`: None
+- `learning_rate`: 3e-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`: 20
+- `max_steps`: -1
+- `lr_scheduler_type`: linear
+- `lr_scheduler_kwargs`: {}
+- `warmup_ratio`: 0.1
+- `warmup_steps`: 0
+- `log_level`: passive
+- `log_level_replica`: warning
+- `log_on_each_node`: True
+- `logging_nan_inf_filter`: True
+- `save_safetensors`: True
+- `save_on_each_node`: False
+- `save_only_model`: False
+- `restore_callback_states_from_checkpoint`: False
+- `no_cuda`: False
+- `use_cpu`: False
+- `use_mps_device`: False
+- `seed`: 42
+- `data_seed`: None
+- `jit_mode_eval`: False
+- `use_ipex`: False
+- `bf16`: False
+- `fp16`: False
+- `fp16_opt_level`: O1
+- `half_precision_backend`: auto
+- `bf16_full_eval`: False
+- `fp16_full_eval`: False
+- `tf32`: None
+- `local_rank`: 0
+- `ddp_backend`: None
+- `tpu_num_cores`: None
+- `tpu_metrics_debug`: False
+- `debug`: []
+- `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`: False
+- `ignore_data_skip`: False
+- `fsdp`: []
+- `fsdp_min_num_params`: 0
+- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
+- `fsdp_transformer_layer_cls_to_wrap`: None
+- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
+- `deepspeed`: None
+- `label_smoothing_factor`: 0.0
+- `optim`: adamw_torch
+- `optim_args`: None
+- `adafactor`: False
+- `group_by_length`: False
+- `length_column_name`: length
+- `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`: False
+- `resume_from_checkpoint`: None
+- `hub_model_id`: None
+- `hub_strategy`: every_save
+- `hub_private_repo`: None
+- `hub_always_push`: False
+- `gradient_checkpointing`: False
+- `gradient_checkpointing_kwargs`: None
+- `include_inputs_for_metrics`: False
+- `include_for_metrics`: []
+- `eval_do_concat_batches`: True
+- `fp16_backend`: auto
+- `push_to_hub_model_id`: None
+- `push_to_hub_organization`: None
+- `mp_parameters`:
+- `auto_find_batch_size`: False
+- `full_determinism`: False
+- `torchdynamo`: None
+- `ray_scope`: last
+- `ddp_timeout`: 1800
+- `torch_compile`: False
+- `torch_compile_backend`: None
+- `torch_compile_mode`: None
+- `dispatch_batches`: None
+- `split_batches`: None
+- `include_tokens_per_second`: False
+- `include_num_input_tokens_seen`: False
+- `neftune_noise_alpha`: None
+- `optim_target_modules`: None
+- `batch_eval_metrics`: False
+- `eval_on_start`: False
+- `use_liger_kernel`: False
+- `eval_use_gather_object`: False
+- `average_tokens_across_devices`: False
+- `prompts`: None
+- `batch_sampler`: batch_sampler
+- `multi_dataset_batch_sampler`: proportional
+
+",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "cls_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "eos_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "mask_token": {
+ "content": "",
+ "lstrip": true,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "pad_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "sep_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "unk_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ }
+}
diff --git a/checkpoints/checkpoint-3000/tokenizer.json b/checkpoints/checkpoint-3000/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ed19c669b24d8e44e6e4c98b8ef4bb83d8cab46
--- /dev/null
+++ b/checkpoints/checkpoint-3000/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de6f09c3f9b891e5b98dd3af9463dcab5a97d5265e288271395324a0577e6c05
+size 133
diff --git a/checkpoints/checkpoint-3000/tokenizer_config.json b/checkpoints/checkpoint-3000/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0df9458a41642f96287ae84a989281a3bdcebacf
--- /dev/null
+++ b/checkpoints/checkpoint-3000/tokenizer_config.json
@@ -0,0 +1,56 @@
+{
+ "added_tokens_decoder": {
+ "0": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "1": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "2": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "3": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "250001": {
+ "content": "",
+ "lstrip": true,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ }
+ },
+ "additional_special_tokens": [],
+ "bos_token": "",
+ "clean_up_tokenization_spaces": true,
+ "cls_token": "",
+ "eos_token": "",
+ "extra_special_tokens": {},
+ "mask_token": "",
+ "model_max_length": 512,
+ "pad_token": "",
+ "sep_token": "",
+ "tokenizer_class": "XLMRobertaTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoints/checkpoint-3000/trainer_state.json b/checkpoints/checkpoint-3000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..d60804bef314ab5e2dab5adc804220cb4fdcb83a
--- /dev/null
+++ b/checkpoints/checkpoint-3000/trainer_state.json
@@ -0,0 +1,2293 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 11.148285449490269,
+ "eval_steps": 150,
+ "global_step": 3000,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.03707136237256719,
+ "grad_norm": 53.36411666870117,
+ "learning_rate": 1.1152416356877324e-06,
+ "loss": 10.8165,
+ "step": 10
+ },
+ {
+ "epoch": 0.07414272474513438,
+ "grad_norm": 61.33858871459961,
+ "learning_rate": 2.2304832713754648e-06,
+ "loss": 10.8463,
+ "step": 20
+ },
+ {
+ "epoch": 0.11121408711770157,
+ "grad_norm": 67.16305541992188,
+ "learning_rate": 3.345724907063197e-06,
+ "loss": 10.821,
+ "step": 30
+ },
+ {
+ "epoch": 0.14828544949026876,
+ "grad_norm": 62.08958053588867,
+ "learning_rate": 4.4609665427509296e-06,
+ "loss": 10.4704,
+ "step": 40
+ },
+ {
+ "epoch": 0.18535681186283595,
+ "grad_norm": 53.839927673339844,
+ "learning_rate": 5.576208178438661e-06,
+ "loss": 10.8568,
+ "step": 50
+ },
+ {
+ "epoch": 0.22242817423540315,
+ "grad_norm": 57.02272033691406,
+ "learning_rate": 6.691449814126394e-06,
+ "loss": 10.3005,
+ "step": 60
+ },
+ {
+ "epoch": 0.2594995366079703,
+ "grad_norm": 88.16449737548828,
+ "learning_rate": 7.806691449814127e-06,
+ "loss": 9.6009,
+ "step": 70
+ },
+ {
+ "epoch": 0.2965708989805375,
+ "grad_norm": 84.35252380371094,
+ "learning_rate": 8.921933085501859e-06,
+ "loss": 9.7448,
+ "step": 80
+ },
+ {
+ "epoch": 0.3336422613531047,
+ "grad_norm": 77.93982696533203,
+ "learning_rate": 1.0037174721189593e-05,
+ "loss": 9.5922,
+ "step": 90
+ },
+ {
+ "epoch": 0.3707136237256719,
+ "grad_norm": 56.669166564941406,
+ "learning_rate": 1.1152416356877323e-05,
+ "loss": 10.0919,
+ "step": 100
+ },
+ {
+ "epoch": 0.4077849860982391,
+ "grad_norm": 56.652488708496094,
+ "learning_rate": 1.2267657992565056e-05,
+ "loss": 10.2733,
+ "step": 110
+ },
+ {
+ "epoch": 0.4448563484708063,
+ "grad_norm": 63.472042083740234,
+ "learning_rate": 1.3382899628252788e-05,
+ "loss": 9.4547,
+ "step": 120
+ },
+ {
+ "epoch": 0.4819277108433735,
+ "grad_norm": 78.0159912109375,
+ "learning_rate": 1.4498141263940521e-05,
+ "loss": 9.1738,
+ "step": 130
+ },
+ {
+ "epoch": 0.5189990732159406,
+ "grad_norm": 68.1931381225586,
+ "learning_rate": 1.5613382899628255e-05,
+ "loss": 10.0876,
+ "step": 140
+ },
+ {
+ "epoch": 0.5560704355885079,
+ "grad_norm": 49.315216064453125,
+ "learning_rate": 1.6728624535315986e-05,
+ "loss": 10.3721,
+ "step": 150
+ },
+ {
+ "epoch": 0.5560704355885079,
+ "eval_loss": 2.559400796890259,
+ "eval_runtime": 34.304,
+ "eval_samples_per_second": 47.166,
+ "eval_steps_per_second": 6.763,
+ "step": 150
+ },
+ {
+ "epoch": 0.593141797961075,
+ "grad_norm": 34.1899528503418,
+ "learning_rate": 1.7843866171003718e-05,
+ "loss": 10.4711,
+ "step": 160
+ },
+ {
+ "epoch": 0.6302131603336423,
+ "grad_norm": 37.074134826660156,
+ "learning_rate": 1.895910780669145e-05,
+ "loss": 9.9383,
+ "step": 170
+ },
+ {
+ "epoch": 0.6672845227062094,
+ "grad_norm": 49.323158264160156,
+ "learning_rate": 2.0074349442379185e-05,
+ "loss": 10.0448,
+ "step": 180
+ },
+ {
+ "epoch": 0.7043558850787767,
+ "grad_norm": 53.07639694213867,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 9.8467,
+ "step": 190
+ },
+ {
+ "epoch": 0.7414272474513438,
+ "grad_norm": 38.345916748046875,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 9.9434,
+ "step": 200
+ },
+ {
+ "epoch": 0.7784986098239111,
+ "grad_norm": 47.55274963378906,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 9.7912,
+ "step": 210
+ },
+ {
+ "epoch": 0.8155699721964782,
+ "grad_norm": 59.17176055908203,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 9.7905,
+ "step": 220
+ },
+ {
+ "epoch": 0.8526413345690455,
+ "grad_norm": 52.57457733154297,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 9.9412,
+ "step": 230
+ },
+ {
+ "epoch": 0.8897126969416126,
+ "grad_norm": 42.251304626464844,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 10.0084,
+ "step": 240
+ },
+ {
+ "epoch": 0.9267840593141798,
+ "grad_norm": 46.63966369628906,
+ "learning_rate": 2.788104089219331e-05,
+ "loss": 10.0139,
+ "step": 250
+ },
+ {
+ "epoch": 0.963855421686747,
+ "grad_norm": 27.772289276123047,
+ "learning_rate": 2.8996282527881043e-05,
+ "loss": 9.3794,
+ "step": 260
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 46.3179931640625,
+ "learning_rate": 2.9987608426270136e-05,
+ "loss": 9.5747,
+ "step": 270
+ },
+ {
+ "epoch": 1.0370713623725671,
+ "grad_norm": 45.746551513671875,
+ "learning_rate": 2.98636926889715e-05,
+ "loss": 9.6763,
+ "step": 280
+ },
+ {
+ "epoch": 1.0741427247451343,
+ "grad_norm": 49.50416564941406,
+ "learning_rate": 2.9739776951672864e-05,
+ "loss": 10.4789,
+ "step": 290
+ },
+ {
+ "epoch": 1.1112140871177016,
+ "grad_norm": 43.72733688354492,
+ "learning_rate": 2.9615861214374226e-05,
+ "loss": 9.7244,
+ "step": 300
+ },
+ {
+ "epoch": 1.1112140871177016,
+ "eval_loss": 2.6879734992980957,
+ "eval_runtime": 34.2522,
+ "eval_samples_per_second": 47.238,
+ "eval_steps_per_second": 6.773,
+ "step": 300
+ },
+ {
+ "epoch": 1.1482854494902688,
+ "grad_norm": 48.68376541137695,
+ "learning_rate": 2.9491945477075588e-05,
+ "loss": 9.9421,
+ "step": 310
+ },
+ {
+ "epoch": 1.185356811862836,
+ "grad_norm": 44.014747619628906,
+ "learning_rate": 2.9368029739776953e-05,
+ "loss": 9.5022,
+ "step": 320
+ },
+ {
+ "epoch": 1.222428174235403,
+ "grad_norm": 49.858642578125,
+ "learning_rate": 2.9244114002478315e-05,
+ "loss": 9.6801,
+ "step": 330
+ },
+ {
+ "epoch": 1.2594995366079704,
+ "grad_norm": 49.97459411621094,
+ "learning_rate": 2.9120198265179677e-05,
+ "loss": 9.25,
+ "step": 340
+ },
+ {
+ "epoch": 1.2965708989805376,
+ "grad_norm": 40.38698196411133,
+ "learning_rate": 2.8996282527881043e-05,
+ "loss": 10.2879,
+ "step": 350
+ },
+ {
+ "epoch": 1.3336422613531047,
+ "grad_norm": 29.797264099121094,
+ "learning_rate": 2.8872366790582405e-05,
+ "loss": 9.97,
+ "step": 360
+ },
+ {
+ "epoch": 1.3707136237256718,
+ "grad_norm": 43.83412551879883,
+ "learning_rate": 2.874845105328377e-05,
+ "loss": 9.3751,
+ "step": 370
+ },
+ {
+ "epoch": 1.407784986098239,
+ "grad_norm": 42.754154205322266,
+ "learning_rate": 2.8624535315985132e-05,
+ "loss": 9.8947,
+ "step": 380
+ },
+ {
+ "epoch": 1.4448563484708064,
+ "grad_norm": 50.08184051513672,
+ "learning_rate": 2.8500619578686494e-05,
+ "loss": 9.4987,
+ "step": 390
+ },
+ {
+ "epoch": 1.4819277108433735,
+ "grad_norm": 41.81029510498047,
+ "learning_rate": 2.8376703841387856e-05,
+ "loss": 10.1515,
+ "step": 400
+ },
+ {
+ "epoch": 1.5189990732159406,
+ "grad_norm": 44.03816223144531,
+ "learning_rate": 2.8252788104089218e-05,
+ "loss": 10.0098,
+ "step": 410
+ },
+ {
+ "epoch": 1.556070435588508,
+ "grad_norm": 42.871856689453125,
+ "learning_rate": 2.8128872366790583e-05,
+ "loss": 9.5566,
+ "step": 420
+ },
+ {
+ "epoch": 1.5931417979610751,
+ "grad_norm": 44.18041229248047,
+ "learning_rate": 2.8004956629491945e-05,
+ "loss": 10.1925,
+ "step": 430
+ },
+ {
+ "epoch": 1.6302131603336423,
+ "grad_norm": 30.012725830078125,
+ "learning_rate": 2.788104089219331e-05,
+ "loss": 9.7408,
+ "step": 440
+ },
+ {
+ "epoch": 1.6672845227062094,
+ "grad_norm": 38.104827880859375,
+ "learning_rate": 2.7757125154894673e-05,
+ "loss": 9.6873,
+ "step": 450
+ },
+ {
+ "epoch": 1.6672845227062094,
+ "eval_loss": 2.5061004161834717,
+ "eval_runtime": 34.2772,
+ "eval_samples_per_second": 47.203,
+ "eval_steps_per_second": 6.768,
+ "step": 450
+ },
+ {
+ "epoch": 1.7043558850787766,
+ "grad_norm": 42.93535232543945,
+ "learning_rate": 2.7633209417596035e-05,
+ "loss": 9.1932,
+ "step": 460
+ },
+ {
+ "epoch": 1.7414272474513437,
+ "grad_norm": 63.935279846191406,
+ "learning_rate": 2.75092936802974e-05,
+ "loss": 9.9204,
+ "step": 470
+ },
+ {
+ "epoch": 1.778498609823911,
+ "grad_norm": 30.906002044677734,
+ "learning_rate": 2.7385377942998762e-05,
+ "loss": 9.5151,
+ "step": 480
+ },
+ {
+ "epoch": 1.8155699721964782,
+ "grad_norm": 48.05679702758789,
+ "learning_rate": 2.7261462205700128e-05,
+ "loss": 9.8418,
+ "step": 490
+ },
+ {
+ "epoch": 1.8526413345690456,
+ "grad_norm": 56.36283874511719,
+ "learning_rate": 2.7137546468401486e-05,
+ "loss": 9.6943,
+ "step": 500
+ },
+ {
+ "epoch": 1.8897126969416127,
+ "grad_norm": 47.22933578491211,
+ "learning_rate": 2.7013630731102848e-05,
+ "loss": 9.6948,
+ "step": 510
+ },
+ {
+ "epoch": 1.9267840593141798,
+ "grad_norm": 36.110008239746094,
+ "learning_rate": 2.6889714993804214e-05,
+ "loss": 9.2274,
+ "step": 520
+ },
+ {
+ "epoch": 1.963855421686747,
+ "grad_norm": 37.13912582397461,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 10.138,
+ "step": 530
+ },
+ {
+ "epoch": 2.0,
+ "grad_norm": 51.37559509277344,
+ "learning_rate": 2.664188351920694e-05,
+ "loss": 8.7934,
+ "step": 540
+ },
+ {
+ "epoch": 2.037071362372567,
+ "grad_norm": 42.32170104980469,
+ "learning_rate": 2.6517967781908303e-05,
+ "loss": 8.8304,
+ "step": 550
+ },
+ {
+ "epoch": 2.0741427247451343,
+ "grad_norm": 42.406219482421875,
+ "learning_rate": 2.6394052044609665e-05,
+ "loss": 8.9593,
+ "step": 560
+ },
+ {
+ "epoch": 2.1112140871177014,
+ "grad_norm": 52.097023010253906,
+ "learning_rate": 2.627013630731103e-05,
+ "loss": 9.0051,
+ "step": 570
+ },
+ {
+ "epoch": 2.1482854494902686,
+ "grad_norm": 83.86861419677734,
+ "learning_rate": 2.6146220570012392e-05,
+ "loss": 9.2829,
+ "step": 580
+ },
+ {
+ "epoch": 2.185356811862836,
+ "grad_norm": 38.229949951171875,
+ "learning_rate": 2.6022304832713758e-05,
+ "loss": 9.785,
+ "step": 590
+ },
+ {
+ "epoch": 2.2224281742354033,
+ "grad_norm": 50.322513580322266,
+ "learning_rate": 2.5898389095415116e-05,
+ "loss": 9.6401,
+ "step": 600
+ },
+ {
+ "epoch": 2.2224281742354033,
+ "eval_loss": 2.448084592819214,
+ "eval_runtime": 34.2525,
+ "eval_samples_per_second": 47.237,
+ "eval_steps_per_second": 6.773,
+ "step": 600
+ },
+ {
+ "epoch": 2.266913809082484,
+ "grad_norm": 39.366676330566406,
+ "learning_rate": 2.5774473358116482e-05,
+ "loss": 8.7059,
+ "step": 610
+ },
+ {
+ "epoch": 2.303985171455051,
+ "grad_norm": 41.00213623046875,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 9.1285,
+ "step": 620
+ },
+ {
+ "epoch": 2.341056533827618,
+ "grad_norm": 63.442169189453125,
+ "learning_rate": 2.5526641883519206e-05,
+ "loss": 9.3277,
+ "step": 630
+ },
+ {
+ "epoch": 2.3781278962001853,
+ "grad_norm": 51.54194641113281,
+ "learning_rate": 2.540272614622057e-05,
+ "loss": 9.1577,
+ "step": 640
+ },
+ {
+ "epoch": 2.4151992585727524,
+ "grad_norm": 57.583885192871094,
+ "learning_rate": 2.5278810408921933e-05,
+ "loss": 9.6032,
+ "step": 650
+ },
+ {
+ "epoch": 2.4522706209453196,
+ "grad_norm": 60.13545227050781,
+ "learning_rate": 2.51548946716233e-05,
+ "loss": 8.5451,
+ "step": 660
+ },
+ {
+ "epoch": 2.489341983317887,
+ "grad_norm": 139.85794067382812,
+ "learning_rate": 2.503097893432466e-05,
+ "loss": 8.6601,
+ "step": 670
+ },
+ {
+ "epoch": 2.5264133456904543,
+ "grad_norm": 160.73171997070312,
+ "learning_rate": 2.4907063197026023e-05,
+ "loss": 9.522,
+ "step": 680
+ },
+ {
+ "epoch": 2.5634847080630214,
+ "grad_norm": 190.7194366455078,
+ "learning_rate": 2.4783147459727385e-05,
+ "loss": 10.0619,
+ "step": 690
+ },
+ {
+ "epoch": 2.6005560704355886,
+ "grad_norm": 172.5655517578125,
+ "learning_rate": 2.4659231722428747e-05,
+ "loss": 9.6444,
+ "step": 700
+ },
+ {
+ "epoch": 2.6376274328081557,
+ "grad_norm": 280.9183044433594,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 10.7593,
+ "step": 710
+ },
+ {
+ "epoch": 2.674698795180723,
+ "grad_norm": 118.03450775146484,
+ "learning_rate": 2.4411400247831474e-05,
+ "loss": 10.4432,
+ "step": 720
+ },
+ {
+ "epoch": 2.71177015755329,
+ "grad_norm": 332.662353515625,
+ "learning_rate": 2.428748451053284e-05,
+ "loss": 9.6033,
+ "step": 730
+ },
+ {
+ "epoch": 2.748841519925857,
+ "grad_norm": 129.18788146972656,
+ "learning_rate": 2.41635687732342e-05,
+ "loss": 9.108,
+ "step": 740
+ },
+ {
+ "epoch": 2.7859128822984243,
+ "grad_norm": 350.5856628417969,
+ "learning_rate": 2.4039653035935564e-05,
+ "loss": 9.884,
+ "step": 750
+ },
+ {
+ "epoch": 2.7859128822984243,
+ "eval_loss": 2.599517583847046,
+ "eval_runtime": 34.0592,
+ "eval_samples_per_second": 47.506,
+ "eval_steps_per_second": 6.812,
+ "step": 750
+ },
+ {
+ "epoch": 2.822984244670992,
+ "grad_norm": 285.14862060546875,
+ "learning_rate": 2.391573729863693e-05,
+ "loss": 11.3952,
+ "step": 760
+ },
+ {
+ "epoch": 2.860055607043559,
+ "grad_norm": 185.5862579345703,
+ "learning_rate": 2.379182156133829e-05,
+ "loss": 10.6798,
+ "step": 770
+ },
+ {
+ "epoch": 2.897126969416126,
+ "grad_norm": 77.20765686035156,
+ "learning_rate": 2.3667905824039656e-05,
+ "loss": 9.9795,
+ "step": 780
+ },
+ {
+ "epoch": 2.9341983317886933,
+ "grad_norm": 88.75109100341797,
+ "learning_rate": 2.3543990086741015e-05,
+ "loss": 9.3694,
+ "step": 790
+ },
+ {
+ "epoch": 2.9712696941612604,
+ "grad_norm": 53.416690826416016,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 8.4631,
+ "step": 800
+ },
+ {
+ "epoch": 3.0111214087117704,
+ "grad_norm": 72.12966918945312,
+ "learning_rate": 2.3296158612143742e-05,
+ "loss": 9.3903,
+ "step": 810
+ },
+ {
+ "epoch": 3.0481927710843375,
+ "grad_norm": 159.3909912109375,
+ "learning_rate": 2.3172242874845104e-05,
+ "loss": 8.9608,
+ "step": 820
+ },
+ {
+ "epoch": 3.0852641334569046,
+ "grad_norm": 161.03713989257812,
+ "learning_rate": 2.304832713754647e-05,
+ "loss": 9.623,
+ "step": 830
+ },
+ {
+ "epoch": 3.122335495829472,
+ "grad_norm": 136.50045776367188,
+ "learning_rate": 2.2924411400247832e-05,
+ "loss": 8.5958,
+ "step": 840
+ },
+ {
+ "epoch": 3.159406858202039,
+ "grad_norm": 60.785926818847656,
+ "learning_rate": 2.2800495662949197e-05,
+ "loss": 8.6278,
+ "step": 850
+ },
+ {
+ "epoch": 3.196478220574606,
+ "grad_norm": 115.2509536743164,
+ "learning_rate": 2.267657992565056e-05,
+ "loss": 8.661,
+ "step": 860
+ },
+ {
+ "epoch": 3.233549582947173,
+ "grad_norm": 144.38912963867188,
+ "learning_rate": 2.255266418835192e-05,
+ "loss": 9.5836,
+ "step": 870
+ },
+ {
+ "epoch": 3.2706209453197403,
+ "grad_norm": 239.74520874023438,
+ "learning_rate": 2.2428748451053287e-05,
+ "loss": 8.6221,
+ "step": 880
+ },
+ {
+ "epoch": 3.3076923076923075,
+ "grad_norm": 215.12789916992188,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 9.1995,
+ "step": 890
+ },
+ {
+ "epoch": 3.344763670064875,
+ "grad_norm": 113.77183532714844,
+ "learning_rate": 2.218091697645601e-05,
+ "loss": 8.6909,
+ "step": 900
+ },
+ {
+ "epoch": 3.344763670064875,
+ "eval_loss": 2.4246325492858887,
+ "eval_runtime": 34.1351,
+ "eval_samples_per_second": 47.4,
+ "eval_steps_per_second": 6.797,
+ "step": 900
+ },
+ {
+ "epoch": 3.381835032437442,
+ "grad_norm": 175.02731323242188,
+ "learning_rate": 2.2057001239157373e-05,
+ "loss": 8.7632,
+ "step": 910
+ },
+ {
+ "epoch": 3.4189063948100094,
+ "grad_norm": 142.76580810546875,
+ "learning_rate": 2.1933085501858735e-05,
+ "loss": 8.5604,
+ "step": 920
+ },
+ {
+ "epoch": 3.4559777571825765,
+ "grad_norm": 163.0391387939453,
+ "learning_rate": 2.18091697645601e-05,
+ "loss": 9.3871,
+ "step": 930
+ },
+ {
+ "epoch": 3.4930491195551436,
+ "grad_norm": 386.4294738769531,
+ "learning_rate": 2.1685254027261462e-05,
+ "loss": 9.6445,
+ "step": 940
+ },
+ {
+ "epoch": 3.5301204819277108,
+ "grad_norm": 86.92581939697266,
+ "learning_rate": 2.1561338289962827e-05,
+ "loss": 9.635,
+ "step": 950
+ },
+ {
+ "epoch": 3.567191844300278,
+ "grad_norm": 80.94845581054688,
+ "learning_rate": 2.143742255266419e-05,
+ "loss": 8.8689,
+ "step": 960
+ },
+ {
+ "epoch": 3.6042632066728455,
+ "grad_norm": 69.99708557128906,
+ "learning_rate": 2.1313506815365555e-05,
+ "loss": 9.1411,
+ "step": 970
+ },
+ {
+ "epoch": 3.641334569045412,
+ "grad_norm": 80.53229522705078,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 8.7552,
+ "step": 980
+ },
+ {
+ "epoch": 3.67840593141798,
+ "grad_norm": 53.3475341796875,
+ "learning_rate": 2.1065675340768275e-05,
+ "loss": 8.5697,
+ "step": 990
+ },
+ {
+ "epoch": 3.715477293790547,
+ "grad_norm": 54.19875717163086,
+ "learning_rate": 2.094175960346964e-05,
+ "loss": 8.9008,
+ "step": 1000
+ },
+ {
+ "epoch": 3.752548656163114,
+ "grad_norm": 52.96628952026367,
+ "learning_rate": 2.0817843866171003e-05,
+ "loss": 8.9415,
+ "step": 1010
+ },
+ {
+ "epoch": 3.789620018535681,
+ "grad_norm": 64.08312225341797,
+ "learning_rate": 2.0693928128872368e-05,
+ "loss": 8.1303,
+ "step": 1020
+ },
+ {
+ "epoch": 3.8266913809082483,
+ "grad_norm": 68.22163391113281,
+ "learning_rate": 2.057001239157373e-05,
+ "loss": 8.241,
+ "step": 1030
+ },
+ {
+ "epoch": 3.8637627432808155,
+ "grad_norm": 66.70902252197266,
+ "learning_rate": 2.0446096654275092e-05,
+ "loss": 8.6969,
+ "step": 1040
+ },
+ {
+ "epoch": 3.9008341056533826,
+ "grad_norm": 49.94658279418945,
+ "learning_rate": 2.0322180916976458e-05,
+ "loss": 9.8392,
+ "step": 1050
+ },
+ {
+ "epoch": 3.9008341056533826,
+ "eval_loss": 2.533297300338745,
+ "eval_runtime": 34.1106,
+ "eval_samples_per_second": 47.434,
+ "eval_steps_per_second": 6.801,
+ "step": 1050
+ },
+ {
+ "epoch": 3.93790546802595,
+ "grad_norm": 73.9527816772461,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 8.5934,
+ "step": 1060
+ },
+ {
+ "epoch": 3.974976830398517,
+ "grad_norm": 76.90135955810547,
+ "learning_rate": 2.6703841387856257e-05,
+ "loss": 8.006,
+ "step": 1070
+ },
+ {
+ "epoch": 4.014828544949027,
+ "grad_norm": 85.86236572265625,
+ "learning_rate": 2.664188351920694e-05,
+ "loss": 9.0081,
+ "step": 1080
+ },
+ {
+ "epoch": 4.051899907321594,
+ "grad_norm": 106.57842254638672,
+ "learning_rate": 2.6579925650557622e-05,
+ "loss": 8.6706,
+ "step": 1090
+ },
+ {
+ "epoch": 4.088971269694161,
+ "grad_norm": 68.10344696044922,
+ "learning_rate": 2.6517967781908303e-05,
+ "loss": 9.6146,
+ "step": 1100
+ },
+ {
+ "epoch": 4.126042632066729,
+ "grad_norm": 67.02091217041016,
+ "learning_rate": 2.6456009913258984e-05,
+ "loss": 9.225,
+ "step": 1110
+ },
+ {
+ "epoch": 4.163113994439295,
+ "grad_norm": 67.66644287109375,
+ "learning_rate": 2.6394052044609665e-05,
+ "loss": 8.7522,
+ "step": 1120
+ },
+ {
+ "epoch": 4.200185356811863,
+ "grad_norm": 59.26814651489258,
+ "learning_rate": 2.633209417596035e-05,
+ "loss": 9.0221,
+ "step": 1130
+ },
+ {
+ "epoch": 4.23725671918443,
+ "grad_norm": 88.07310485839844,
+ "learning_rate": 2.627013630731103e-05,
+ "loss": 9.6458,
+ "step": 1140
+ },
+ {
+ "epoch": 4.274328081556997,
+ "grad_norm": 69.33895874023438,
+ "learning_rate": 2.620817843866171e-05,
+ "loss": 8.7692,
+ "step": 1150
+ },
+ {
+ "epoch": 4.311399443929565,
+ "grad_norm": 43.823951721191406,
+ "learning_rate": 2.6146220570012392e-05,
+ "loss": 9.2874,
+ "step": 1160
+ },
+ {
+ "epoch": 4.348470806302132,
+ "grad_norm": 63.590980529785156,
+ "learning_rate": 2.6084262701363077e-05,
+ "loss": 8.9276,
+ "step": 1170
+ },
+ {
+ "epoch": 4.385542168674699,
+ "grad_norm": 43.4519157409668,
+ "learning_rate": 2.6022304832713758e-05,
+ "loss": 8.7444,
+ "step": 1180
+ },
+ {
+ "epoch": 4.422613531047266,
+ "grad_norm": 72.1529312133789,
+ "learning_rate": 2.5960346964064435e-05,
+ "loss": 8.7265,
+ "step": 1190
+ },
+ {
+ "epoch": 4.459684893419833,
+ "grad_norm": 101.67008972167969,
+ "learning_rate": 2.5898389095415116e-05,
+ "loss": 8.7642,
+ "step": 1200
+ },
+ {
+ "epoch": 4.459684893419833,
+ "eval_loss": 2.6471314430236816,
+ "eval_runtime": 34.2723,
+ "eval_samples_per_second": 47.21,
+ "eval_steps_per_second": 6.769,
+ "step": 1200
+ },
+ {
+ "epoch": 4.4967562557924,
+ "grad_norm": 66.03606414794922,
+ "learning_rate": 2.5836431226765797e-05,
+ "loss": 8.8917,
+ "step": 1210
+ },
+ {
+ "epoch": 4.533827618164968,
+ "grad_norm": 60.01662826538086,
+ "learning_rate": 2.5774473358116482e-05,
+ "loss": 9.2155,
+ "step": 1220
+ },
+ {
+ "epoch": 4.570898980537534,
+ "grad_norm": 83.86510467529297,
+ "learning_rate": 2.5712515489467163e-05,
+ "loss": 8.6101,
+ "step": 1230
+ },
+ {
+ "epoch": 4.607970342910102,
+ "grad_norm": 64.35486602783203,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 8.9904,
+ "step": 1240
+ },
+ {
+ "epoch": 4.645041705282669,
+ "grad_norm": 60.44017791748047,
+ "learning_rate": 2.5588599752168525e-05,
+ "loss": 9.3272,
+ "step": 1250
+ },
+ {
+ "epoch": 4.682113067655236,
+ "grad_norm": 56.501102447509766,
+ "learning_rate": 2.5526641883519206e-05,
+ "loss": 7.9367,
+ "step": 1260
+ },
+ {
+ "epoch": 4.719184430027804,
+ "grad_norm": 65.01017761230469,
+ "learning_rate": 2.546468401486989e-05,
+ "loss": 8.5891,
+ "step": 1270
+ },
+ {
+ "epoch": 4.7562557924003706,
+ "grad_norm": 63.42623519897461,
+ "learning_rate": 2.540272614622057e-05,
+ "loss": 8.6286,
+ "step": 1280
+ },
+ {
+ "epoch": 4.793327154772938,
+ "grad_norm": 58.96747589111328,
+ "learning_rate": 2.5340768277571252e-05,
+ "loss": 7.9982,
+ "step": 1290
+ },
+ {
+ "epoch": 4.830398517145505,
+ "grad_norm": 96.77368927001953,
+ "learning_rate": 2.5278810408921933e-05,
+ "loss": 7.5587,
+ "step": 1300
+ },
+ {
+ "epoch": 4.867469879518072,
+ "grad_norm": 85.92269897460938,
+ "learning_rate": 2.5216852540272614e-05,
+ "loss": 7.9405,
+ "step": 1310
+ },
+ {
+ "epoch": 4.904541241890639,
+ "grad_norm": 86.37085723876953,
+ "learning_rate": 2.51548946716233e-05,
+ "loss": 9.7092,
+ "step": 1320
+ },
+ {
+ "epoch": 4.941612604263207,
+ "grad_norm": 65.88274383544922,
+ "learning_rate": 2.509293680297398e-05,
+ "loss": 8.1475,
+ "step": 1330
+ },
+ {
+ "epoch": 4.978683966635774,
+ "grad_norm": 53.252742767333984,
+ "learning_rate": 2.503097893432466e-05,
+ "loss": 9.3603,
+ "step": 1340
+ },
+ {
+ "epoch": 5.014828544949027,
+ "grad_norm": 182.35781860351562,
+ "learning_rate": 2.4969021065675342e-05,
+ "loss": 7.6621,
+ "step": 1350
+ },
+ {
+ "epoch": 5.014828544949027,
+ "eval_loss": 2.8308892250061035,
+ "eval_runtime": 34.1063,
+ "eval_samples_per_second": 47.44,
+ "eval_steps_per_second": 6.802,
+ "step": 1350
+ },
+ {
+ "epoch": 5.051899907321594,
+ "grad_norm": 324.6659240722656,
+ "learning_rate": 2.4907063197026023e-05,
+ "loss": 9.2301,
+ "step": 1360
+ },
+ {
+ "epoch": 5.088971269694161,
+ "grad_norm": 4839.4453125,
+ "learning_rate": 2.4845105328376707e-05,
+ "loss": 9.7789,
+ "step": 1370
+ },
+ {
+ "epoch": 5.126042632066729,
+ "grad_norm": 408588.40625,
+ "learning_rate": 2.4783147459727385e-05,
+ "loss": 9.5359,
+ "step": 1380
+ },
+ {
+ "epoch": 5.163113994439295,
+ "grad_norm": 932.5144653320312,
+ "learning_rate": 2.4721189591078066e-05,
+ "loss": 10.8065,
+ "step": 1390
+ },
+ {
+ "epoch": 5.200185356811863,
+ "grad_norm": 148.0634765625,
+ "learning_rate": 2.4659231722428747e-05,
+ "loss": 10.0149,
+ "step": 1400
+ },
+ {
+ "epoch": 5.23725671918443,
+ "grad_norm": 1214.7279052734375,
+ "learning_rate": 2.459727385377943e-05,
+ "loss": 10.2582,
+ "step": 1410
+ },
+ {
+ "epoch": 5.274328081556997,
+ "grad_norm": 188.14292907714844,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 10.16,
+ "step": 1420
+ },
+ {
+ "epoch": 5.311399443929565,
+ "grad_norm": 388.7134094238281,
+ "learning_rate": 2.4473358116480793e-05,
+ "loss": 10.0763,
+ "step": 1430
+ },
+ {
+ "epoch": 5.348470806302132,
+ "grad_norm": 270.61279296875,
+ "learning_rate": 2.4411400247831474e-05,
+ "loss": 9.5737,
+ "step": 1440
+ },
+ {
+ "epoch": 5.385542168674699,
+ "grad_norm": 1637.854736328125,
+ "learning_rate": 2.4349442379182155e-05,
+ "loss": 10.4816,
+ "step": 1450
+ },
+ {
+ "epoch": 5.422613531047266,
+ "grad_norm": 1118.7012939453125,
+ "learning_rate": 2.428748451053284e-05,
+ "loss": 8.6687,
+ "step": 1460
+ },
+ {
+ "epoch": 5.459684893419833,
+ "grad_norm": 788.9707641601562,
+ "learning_rate": 2.422552664188352e-05,
+ "loss": 8.4066,
+ "step": 1470
+ },
+ {
+ "epoch": 5.4967562557924,
+ "grad_norm": 185.22950744628906,
+ "learning_rate": 2.41635687732342e-05,
+ "loss": 9.386,
+ "step": 1480
+ },
+ {
+ "epoch": 5.533827618164968,
+ "grad_norm": 257.726806640625,
+ "learning_rate": 2.4101610904584883e-05,
+ "loss": 8.3911,
+ "step": 1490
+ },
+ {
+ "epoch": 5.570898980537534,
+ "grad_norm": 867.30908203125,
+ "learning_rate": 2.4039653035935564e-05,
+ "loss": 8.8025,
+ "step": 1500
+ },
+ {
+ "epoch": 5.570898980537534,
+ "eval_loss": 2.5407650470733643,
+ "eval_runtime": 34.0934,
+ "eval_samples_per_second": 47.458,
+ "eval_steps_per_second": 6.805,
+ "step": 1500
+ },
+ {
+ "epoch": 5.607970342910102,
+ "grad_norm": 68.35711669921875,
+ "learning_rate": 2.3977695167286248e-05,
+ "loss": 8.7939,
+ "step": 1510
+ },
+ {
+ "epoch": 5.645041705282669,
+ "grad_norm": 1419.1650390625,
+ "learning_rate": 2.391573729863693e-05,
+ "loss": 9.0903,
+ "step": 1520
+ },
+ {
+ "epoch": 5.682113067655236,
+ "grad_norm": 60.88732147216797,
+ "learning_rate": 2.385377942998761e-05,
+ "loss": 8.9878,
+ "step": 1530
+ },
+ {
+ "epoch": 5.719184430027804,
+ "grad_norm": 61.50148010253906,
+ "learning_rate": 2.379182156133829e-05,
+ "loss": 8.8642,
+ "step": 1540
+ },
+ {
+ "epoch": 5.7562557924003706,
+ "grad_norm": 73.9307861328125,
+ "learning_rate": 2.3729863692688972e-05,
+ "loss": 8.8625,
+ "step": 1550
+ },
+ {
+ "epoch": 5.793327154772938,
+ "grad_norm": 82.28329467773438,
+ "learning_rate": 2.3667905824039656e-05,
+ "loss": 8.4105,
+ "step": 1560
+ },
+ {
+ "epoch": 5.830398517145505,
+ "grad_norm": 83.54141235351562,
+ "learning_rate": 2.3605947955390337e-05,
+ "loss": 9.0163,
+ "step": 1570
+ },
+ {
+ "epoch": 5.867469879518072,
+ "grad_norm": 66.59397888183594,
+ "learning_rate": 2.3543990086741015e-05,
+ "loss": 8.8947,
+ "step": 1580
+ },
+ {
+ "epoch": 5.904541241890639,
+ "grad_norm": 55.8841667175293,
+ "learning_rate": 2.3482032218091696e-05,
+ "loss": 8.5647,
+ "step": 1590
+ },
+ {
+ "epoch": 5.941612604263207,
+ "grad_norm": 74.56892395019531,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 7.7047,
+ "step": 1600
+ },
+ {
+ "epoch": 5.978683966635774,
+ "grad_norm": 106.54405212402344,
+ "learning_rate": 2.335811648079306e-05,
+ "loss": 8.1484,
+ "step": 1610
+ },
+ {
+ "epoch": 6.014828544949027,
+ "grad_norm": 107.8216781616211,
+ "learning_rate": 2.3296158612143742e-05,
+ "loss": 8.4079,
+ "step": 1620
+ },
+ {
+ "epoch": 6.051899907321594,
+ "grad_norm": 100.98133087158203,
+ "learning_rate": 2.3234200743494423e-05,
+ "loss": 8.5027,
+ "step": 1630
+ },
+ {
+ "epoch": 6.088971269694161,
+ "grad_norm": 67.31327819824219,
+ "learning_rate": 2.3172242874845104e-05,
+ "loss": 8.1805,
+ "step": 1640
+ },
+ {
+ "epoch": 6.126042632066729,
+ "grad_norm": 61.10900115966797,
+ "learning_rate": 2.311028500619579e-05,
+ "loss": 8.4519,
+ "step": 1650
+ },
+ {
+ "epoch": 6.126042632066729,
+ "eval_loss": 2.590090036392212,
+ "eval_runtime": 34.1392,
+ "eval_samples_per_second": 47.394,
+ "eval_steps_per_second": 6.796,
+ "step": 1650
+ },
+ {
+ "epoch": 6.163113994439295,
+ "grad_norm": 60.91183090209961,
+ "learning_rate": 2.304832713754647e-05,
+ "loss": 9.062,
+ "step": 1660
+ },
+ {
+ "epoch": 6.200185356811863,
+ "grad_norm": 76.2873306274414,
+ "learning_rate": 2.298636926889715e-05,
+ "loss": 8.8499,
+ "step": 1670
+ },
+ {
+ "epoch": 6.23725671918443,
+ "grad_norm": 74.2361068725586,
+ "learning_rate": 2.2924411400247832e-05,
+ "loss": 8.6576,
+ "step": 1680
+ },
+ {
+ "epoch": 6.274328081556997,
+ "grad_norm": 84.99951934814453,
+ "learning_rate": 2.2862453531598513e-05,
+ "loss": 8.4652,
+ "step": 1690
+ },
+ {
+ "epoch": 6.311399443929565,
+ "grad_norm": 86.40947723388672,
+ "learning_rate": 2.2800495662949197e-05,
+ "loss": 9.0782,
+ "step": 1700
+ },
+ {
+ "epoch": 6.348470806302132,
+ "grad_norm": 75.45568084716797,
+ "learning_rate": 2.2738537794299878e-05,
+ "loss": 8.1532,
+ "step": 1710
+ },
+ {
+ "epoch": 6.385542168674699,
+ "grad_norm": 86.65908813476562,
+ "learning_rate": 2.267657992565056e-05,
+ "loss": 8.5185,
+ "step": 1720
+ },
+ {
+ "epoch": 6.422613531047266,
+ "grad_norm": 62.620758056640625,
+ "learning_rate": 2.261462205700124e-05,
+ "loss": 9.5908,
+ "step": 1730
+ },
+ {
+ "epoch": 6.459684893419833,
+ "grad_norm": 73.63095092773438,
+ "learning_rate": 2.255266418835192e-05,
+ "loss": 8.4188,
+ "step": 1740
+ },
+ {
+ "epoch": 6.4967562557924,
+ "grad_norm": 55.175045013427734,
+ "learning_rate": 2.2490706319702606e-05,
+ "loss": 8.1885,
+ "step": 1750
+ },
+ {
+ "epoch": 6.533827618164968,
+ "grad_norm": 47.647621154785156,
+ "learning_rate": 2.2428748451053287e-05,
+ "loss": 8.7666,
+ "step": 1760
+ },
+ {
+ "epoch": 6.570898980537534,
+ "grad_norm": 191.0419158935547,
+ "learning_rate": 2.2366790582403964e-05,
+ "loss": 8.6105,
+ "step": 1770
+ },
+ {
+ "epoch": 6.607970342910102,
+ "grad_norm": 57.81939697265625,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 8.664,
+ "step": 1780
+ },
+ {
+ "epoch": 6.645041705282669,
+ "grad_norm": 85.04927825927734,
+ "learning_rate": 2.2242874845105326e-05,
+ "loss": 8.5294,
+ "step": 1790
+ },
+ {
+ "epoch": 6.682113067655236,
+ "grad_norm": 55.604122161865234,
+ "learning_rate": 2.218091697645601e-05,
+ "loss": 9.1857,
+ "step": 1800
+ },
+ {
+ "epoch": 6.682113067655236,
+ "eval_loss": 2.497358560562134,
+ "eval_runtime": 34.0441,
+ "eval_samples_per_second": 47.527,
+ "eval_steps_per_second": 6.815,
+ "step": 1800
+ },
+ {
+ "epoch": 6.719184430027804,
+ "grad_norm": 75.21812438964844,
+ "learning_rate": 2.211895910780669e-05,
+ "loss": 8.7053,
+ "step": 1810
+ },
+ {
+ "epoch": 6.7562557924003706,
+ "grad_norm": 85.76676940917969,
+ "learning_rate": 2.2057001239157373e-05,
+ "loss": 8.1428,
+ "step": 1820
+ },
+ {
+ "epoch": 6.793327154772938,
+ "grad_norm": 69.85575866699219,
+ "learning_rate": 2.1995043370508054e-05,
+ "loss": 8.4988,
+ "step": 1830
+ },
+ {
+ "epoch": 6.830398517145505,
+ "grad_norm": 66.85325622558594,
+ "learning_rate": 2.1933085501858735e-05,
+ "loss": 8.4147,
+ "step": 1840
+ },
+ {
+ "epoch": 6.867469879518072,
+ "grad_norm": 60.645294189453125,
+ "learning_rate": 2.187112763320942e-05,
+ "loss": 9.069,
+ "step": 1850
+ },
+ {
+ "epoch": 6.904541241890639,
+ "grad_norm": 77.59921264648438,
+ "learning_rate": 2.18091697645601e-05,
+ "loss": 8.4405,
+ "step": 1860
+ },
+ {
+ "epoch": 6.941612604263207,
+ "grad_norm": 56.27996826171875,
+ "learning_rate": 2.174721189591078e-05,
+ "loss": 9.2157,
+ "step": 1870
+ },
+ {
+ "epoch": 6.978683966635774,
+ "grad_norm": 64.18400573730469,
+ "learning_rate": 2.1685254027261462e-05,
+ "loss": 9.5492,
+ "step": 1880
+ },
+ {
+ "epoch": 7.014828544949027,
+ "grad_norm": 76.78484344482422,
+ "learning_rate": 2.1623296158612143e-05,
+ "loss": 8.1325,
+ "step": 1890
+ },
+ {
+ "epoch": 7.051899907321594,
+ "grad_norm": 77.69871520996094,
+ "learning_rate": 2.1561338289962827e-05,
+ "loss": 8.324,
+ "step": 1900
+ },
+ {
+ "epoch": 7.088971269694161,
+ "grad_norm": 102.43340301513672,
+ "learning_rate": 2.149938042131351e-05,
+ "loss": 7.7097,
+ "step": 1910
+ },
+ {
+ "epoch": 7.126042632066729,
+ "grad_norm": 114.0189208984375,
+ "learning_rate": 2.143742255266419e-05,
+ "loss": 8.0982,
+ "step": 1920
+ },
+ {
+ "epoch": 7.163113994439295,
+ "grad_norm": 66.60838317871094,
+ "learning_rate": 2.137546468401487e-05,
+ "loss": 7.7669,
+ "step": 1930
+ },
+ {
+ "epoch": 7.200185356811863,
+ "grad_norm": 252.72503662109375,
+ "learning_rate": 2.1313506815365555e-05,
+ "loss": 7.809,
+ "step": 1940
+ },
+ {
+ "epoch": 7.23725671918443,
+ "grad_norm": 91.0379409790039,
+ "learning_rate": 2.1251548946716236e-05,
+ "loss": 7.9729,
+ "step": 1950
+ },
+ {
+ "epoch": 7.23725671918443,
+ "eval_loss": 2.610764741897583,
+ "eval_runtime": 34.0774,
+ "eval_samples_per_second": 47.48,
+ "eval_steps_per_second": 6.808,
+ "step": 1950
+ },
+ {
+ "epoch": 7.274328081556997,
+ "grad_norm": 96.09555053710938,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 8.2125,
+ "step": 1960
+ },
+ {
+ "epoch": 7.311399443929565,
+ "grad_norm": 68.51132202148438,
+ "learning_rate": 2.1127633209417594e-05,
+ "loss": 7.7403,
+ "step": 1970
+ },
+ {
+ "epoch": 7.348470806302132,
+ "grad_norm": 94.45308685302734,
+ "learning_rate": 2.1065675340768275e-05,
+ "loss": 7.5494,
+ "step": 1980
+ },
+ {
+ "epoch": 7.385542168674699,
+ "grad_norm": 56.785587310791016,
+ "learning_rate": 2.100371747211896e-05,
+ "loss": 8.2821,
+ "step": 1990
+ },
+ {
+ "epoch": 7.422613531047266,
+ "grad_norm": 91.59173583984375,
+ "learning_rate": 2.094175960346964e-05,
+ "loss": 8.1644,
+ "step": 2000
+ },
+ {
+ "epoch": 7.459684893419833,
+ "grad_norm": 78.57572174072266,
+ "learning_rate": 2.0879801734820322e-05,
+ "loss": 8.1664,
+ "step": 2010
+ },
+ {
+ "epoch": 7.4967562557924,
+ "grad_norm": 71.83062744140625,
+ "learning_rate": 2.0817843866171003e-05,
+ "loss": 8.5876,
+ "step": 2020
+ },
+ {
+ "epoch": 7.533827618164968,
+ "grad_norm": 74.75129699707031,
+ "learning_rate": 2.0755885997521684e-05,
+ "loss": 8.2753,
+ "step": 2030
+ },
+ {
+ "epoch": 7.570898980537534,
+ "grad_norm": 96.26950073242188,
+ "learning_rate": 2.0693928128872368e-05,
+ "loss": 9.2057,
+ "step": 2040
+ },
+ {
+ "epoch": 7.607970342910102,
+ "grad_norm": 96.87144470214844,
+ "learning_rate": 2.063197026022305e-05,
+ "loss": 8.0052,
+ "step": 2050
+ },
+ {
+ "epoch": 7.645041705282669,
+ "grad_norm": 83.44422912597656,
+ "learning_rate": 2.057001239157373e-05,
+ "loss": 8.4954,
+ "step": 2060
+ },
+ {
+ "epoch": 7.682113067655236,
+ "grad_norm": 86.1534194946289,
+ "learning_rate": 2.050805452292441e-05,
+ "loss": 8.0325,
+ "step": 2070
+ },
+ {
+ "epoch": 7.719184430027804,
+ "grad_norm": 89.11744689941406,
+ "learning_rate": 2.0446096654275092e-05,
+ "loss": 8.2934,
+ "step": 2080
+ },
+ {
+ "epoch": 7.7562557924003706,
+ "grad_norm": 60.81666946411133,
+ "learning_rate": 2.0384138785625777e-05,
+ "loss": 9.4019,
+ "step": 2090
+ },
+ {
+ "epoch": 7.793327154772938,
+ "grad_norm": 82.93228912353516,
+ "learning_rate": 2.0322180916976458e-05,
+ "loss": 8.874,
+ "step": 2100
+ },
+ {
+ "epoch": 7.793327154772938,
+ "eval_loss": 2.452894926071167,
+ "eval_runtime": 34.0684,
+ "eval_samples_per_second": 47.493,
+ "eval_steps_per_second": 6.81,
+ "step": 2100
+ },
+ {
+ "epoch": 7.830398517145505,
+ "grad_norm": 62.004817962646484,
+ "learning_rate": 2.026022304832714e-05,
+ "loss": 8.0985,
+ "step": 2110
+ },
+ {
+ "epoch": 7.867469879518072,
+ "grad_norm": 72.79752349853516,
+ "learning_rate": 2.019826517967782e-05,
+ "loss": 8.9063,
+ "step": 2120
+ },
+ {
+ "epoch": 7.904541241890639,
+ "grad_norm": 1507.2855224609375,
+ "learning_rate": 2.01363073110285e-05,
+ "loss": 8.6332,
+ "step": 2130
+ },
+ {
+ "epoch": 7.941612604263207,
+ "grad_norm": 134.11378479003906,
+ "learning_rate": 2.0074349442379185e-05,
+ "loss": 9.4106,
+ "step": 2140
+ },
+ {
+ "epoch": 7.978683966635774,
+ "grad_norm": 72.19046783447266,
+ "learning_rate": 2.0012391573729866e-05,
+ "loss": 8.2304,
+ "step": 2150
+ },
+ {
+ "epoch": 8.014828544949028,
+ "grad_norm": 57.130165100097656,
+ "learning_rate": 1.9950433705080544e-05,
+ "loss": 7.9675,
+ "step": 2160
+ },
+ {
+ "epoch": 8.051899907321594,
+ "grad_norm": 72.58036041259766,
+ "learning_rate": 1.9888475836431225e-05,
+ "loss": 7.602,
+ "step": 2170
+ },
+ {
+ "epoch": 8.088971269694161,
+ "grad_norm": 91.77513885498047,
+ "learning_rate": 1.982651796778191e-05,
+ "loss": 7.9151,
+ "step": 2180
+ },
+ {
+ "epoch": 8.126042632066728,
+ "grad_norm": 111.52696228027344,
+ "learning_rate": 1.976456009913259e-05,
+ "loss": 7.5653,
+ "step": 2190
+ },
+ {
+ "epoch": 8.163113994439296,
+ "grad_norm": 74.9671401977539,
+ "learning_rate": 1.970260223048327e-05,
+ "loss": 7.1567,
+ "step": 2200
+ },
+ {
+ "epoch": 8.200185356811863,
+ "grad_norm": 124.09498596191406,
+ "learning_rate": 1.9640644361833952e-05,
+ "loss": 8.1708,
+ "step": 2210
+ },
+ {
+ "epoch": 8.23725671918443,
+ "grad_norm": 95.21754455566406,
+ "learning_rate": 1.9578686493184633e-05,
+ "loss": 7.0013,
+ "step": 2220
+ },
+ {
+ "epoch": 8.274328081556996,
+ "grad_norm": 135.3839569091797,
+ "learning_rate": 1.9516728624535318e-05,
+ "loss": 7.5999,
+ "step": 2230
+ },
+ {
+ "epoch": 8.311399443929565,
+ "grad_norm": 89.70751190185547,
+ "learning_rate": 1.9454770755886e-05,
+ "loss": 7.2028,
+ "step": 2240
+ },
+ {
+ "epoch": 8.348470806302132,
+ "grad_norm": 70.08137512207031,
+ "learning_rate": 1.939281288723668e-05,
+ "loss": 8.3759,
+ "step": 2250
+ },
+ {
+ "epoch": 8.348470806302132,
+ "eval_loss": 2.746619939804077,
+ "eval_runtime": 34.0831,
+ "eval_samples_per_second": 47.472,
+ "eval_steps_per_second": 6.807,
+ "step": 2250
+ },
+ {
+ "epoch": 8.400370713623726,
+ "grad_norm": 69.07305145263672,
+ "learning_rate": 1.933085501858736e-05,
+ "loss": 8.9083,
+ "step": 2260
+ },
+ {
+ "epoch": 8.437442075996293,
+ "grad_norm": 66.43407440185547,
+ "learning_rate": 1.926889714993804e-05,
+ "loss": 6.9349,
+ "step": 2270
+ },
+ {
+ "epoch": 8.47451343836886,
+ "grad_norm": 167.85784912109375,
+ "learning_rate": 1.9206939281288726e-05,
+ "loss": 7.5041,
+ "step": 2280
+ },
+ {
+ "epoch": 8.511584800741428,
+ "grad_norm": 173.38304138183594,
+ "learning_rate": 1.9144981412639407e-05,
+ "loss": 7.3744,
+ "step": 2290
+ },
+ {
+ "epoch": 8.548656163113995,
+ "grad_norm": 80.71511840820312,
+ "learning_rate": 1.9083023543990088e-05,
+ "loss": 8.6541,
+ "step": 2300
+ },
+ {
+ "epoch": 8.585727525486561,
+ "grad_norm": 82.77088928222656,
+ "learning_rate": 1.902106567534077e-05,
+ "loss": 8.6305,
+ "step": 2310
+ },
+ {
+ "epoch": 8.62279888785913,
+ "grad_norm": 55.3135871887207,
+ "learning_rate": 1.895910780669145e-05,
+ "loss": 8.2577,
+ "step": 2320
+ },
+ {
+ "epoch": 8.659870250231696,
+ "grad_norm": 64.61204528808594,
+ "learning_rate": 1.8897149938042134e-05,
+ "loss": 7.6382,
+ "step": 2330
+ },
+ {
+ "epoch": 8.696941612604263,
+ "grad_norm": 77.9918212890625,
+ "learning_rate": 1.8835192069392815e-05,
+ "loss": 8.114,
+ "step": 2340
+ },
+ {
+ "epoch": 8.73401297497683,
+ "grad_norm": 50.139678955078125,
+ "learning_rate": 1.8773234200743496e-05,
+ "loss": 7.8875,
+ "step": 2350
+ },
+ {
+ "epoch": 8.771084337349398,
+ "grad_norm": 241.43588256835938,
+ "learning_rate": 1.8711276332094174e-05,
+ "loss": 7.0444,
+ "step": 2360
+ },
+ {
+ "epoch": 8.808155699721965,
+ "grad_norm": 133.25645446777344,
+ "learning_rate": 1.8649318463444855e-05,
+ "loss": 7.7393,
+ "step": 2370
+ },
+ {
+ "epoch": 8.845227062094532,
+ "grad_norm": 94.1192855834961,
+ "learning_rate": 1.858736059479554e-05,
+ "loss": 8.8284,
+ "step": 2380
+ },
+ {
+ "epoch": 8.882298424467098,
+ "grad_norm": 83.68805694580078,
+ "learning_rate": 1.852540272614622e-05,
+ "loss": 7.997,
+ "step": 2390
+ },
+ {
+ "epoch": 8.919369786839667,
+ "grad_norm": 67.01426696777344,
+ "learning_rate": 1.84634448574969e-05,
+ "loss": 7.786,
+ "step": 2400
+ },
+ {
+ "epoch": 8.919369786839667,
+ "eval_loss": 2.6791422367095947,
+ "eval_runtime": 34.2417,
+ "eval_samples_per_second": 47.252,
+ "eval_steps_per_second": 6.775,
+ "step": 2400
+ },
+ {
+ "epoch": 8.956441149212234,
+ "grad_norm": 72.36073303222656,
+ "learning_rate": 1.8401486988847582e-05,
+ "loss": 7.6257,
+ "step": 2410
+ },
+ {
+ "epoch": 8.9935125115848,
+ "grad_norm": 70.3509521484375,
+ "learning_rate": 1.8339529120198267e-05,
+ "loss": 7.099,
+ "step": 2420
+ },
+ {
+ "epoch": 9.03336422613531,
+ "grad_norm": 136.34727478027344,
+ "learning_rate": 1.8277571251548948e-05,
+ "loss": 8.041,
+ "step": 2430
+ },
+ {
+ "epoch": 9.070435588507877,
+ "grad_norm": 81.48332214355469,
+ "learning_rate": 1.821561338289963e-05,
+ "loss": 7.8606,
+ "step": 2440
+ },
+ {
+ "epoch": 9.107506950880445,
+ "grad_norm": 93.1724853515625,
+ "learning_rate": 1.815365551425031e-05,
+ "loss": 7.8551,
+ "step": 2450
+ },
+ {
+ "epoch": 9.144578313253012,
+ "grad_norm": 132.82778930664062,
+ "learning_rate": 1.809169764560099e-05,
+ "loss": 7.3977,
+ "step": 2460
+ },
+ {
+ "epoch": 9.181649675625579,
+ "grad_norm": 102.73971557617188,
+ "learning_rate": 1.8029739776951675e-05,
+ "loss": 7.8721,
+ "step": 2470
+ },
+ {
+ "epoch": 9.218721037998147,
+ "grad_norm": 106.7470932006836,
+ "learning_rate": 1.7967781908302356e-05,
+ "loss": 7.5839,
+ "step": 2480
+ },
+ {
+ "epoch": 9.255792400370714,
+ "grad_norm": 84.64981842041016,
+ "learning_rate": 1.7905824039653037e-05,
+ "loss": 7.1823,
+ "step": 2490
+ },
+ {
+ "epoch": 9.29286376274328,
+ "grad_norm": 85.59486389160156,
+ "learning_rate": 1.7843866171003718e-05,
+ "loss": 7.5513,
+ "step": 2500
+ },
+ {
+ "epoch": 9.329935125115847,
+ "grad_norm": 89.53803253173828,
+ "learning_rate": 1.77819083023544e-05,
+ "loss": 8.0879,
+ "step": 2510
+ },
+ {
+ "epoch": 9.367006487488416,
+ "grad_norm": 43.886898040771484,
+ "learning_rate": 1.7719950433705084e-05,
+ "loss": 7.5694,
+ "step": 2520
+ },
+ {
+ "epoch": 9.404077849860982,
+ "grad_norm": 73.4898910522461,
+ "learning_rate": 1.7657992565055765e-05,
+ "loss": 7.3436,
+ "step": 2530
+ },
+ {
+ "epoch": 9.44114921223355,
+ "grad_norm": 100.42314147949219,
+ "learning_rate": 1.7596034696406446e-05,
+ "loss": 6.9425,
+ "step": 2540
+ },
+ {
+ "epoch": 9.478220574606116,
+ "grad_norm": 121.53675079345703,
+ "learning_rate": 1.7534076827757123e-05,
+ "loss": 7.9461,
+ "step": 2550
+ },
+ {
+ "epoch": 9.478220574606116,
+ "eval_loss": 2.6608545780181885,
+ "eval_runtime": 34.2437,
+ "eval_samples_per_second": 47.25,
+ "eval_steps_per_second": 6.775,
+ "step": 2550
+ },
+ {
+ "epoch": 9.515291936978684,
+ "grad_norm": 83.28184509277344,
+ "learning_rate": 1.7472118959107804e-05,
+ "loss": 6.782,
+ "step": 2560
+ },
+ {
+ "epoch": 9.552363299351251,
+ "grad_norm": 81.46753692626953,
+ "learning_rate": 1.741016109045849e-05,
+ "loss": 7.3027,
+ "step": 2570
+ },
+ {
+ "epoch": 9.589434661723818,
+ "grad_norm": 77.35752868652344,
+ "learning_rate": 1.734820322180917e-05,
+ "loss": 7.3348,
+ "step": 2580
+ },
+ {
+ "epoch": 9.626506024096386,
+ "grad_norm": 77.17479705810547,
+ "learning_rate": 1.728624535315985e-05,
+ "loss": 7.7864,
+ "step": 2590
+ },
+ {
+ "epoch": 9.663577386468953,
+ "grad_norm": 103.42965698242188,
+ "learning_rate": 1.722428748451053e-05,
+ "loss": 6.3552,
+ "step": 2600
+ },
+ {
+ "epoch": 9.70064874884152,
+ "grad_norm": 103.3230972290039,
+ "learning_rate": 1.7162329615861213e-05,
+ "loss": 7.151,
+ "step": 2610
+ },
+ {
+ "epoch": 9.737720111214086,
+ "grad_norm": 91.23753356933594,
+ "learning_rate": 1.7100371747211897e-05,
+ "loss": 6.1664,
+ "step": 2620
+ },
+ {
+ "epoch": 9.774791473586655,
+ "grad_norm": 68.76666259765625,
+ "learning_rate": 1.7038413878562578e-05,
+ "loss": 6.0398,
+ "step": 2630
+ },
+ {
+ "epoch": 9.811862835959221,
+ "grad_norm": 112.56900787353516,
+ "learning_rate": 1.697645600991326e-05,
+ "loss": 7.0452,
+ "step": 2640
+ },
+ {
+ "epoch": 9.848934198331788,
+ "grad_norm": 142.0828399658203,
+ "learning_rate": 1.691449814126394e-05,
+ "loss": 7.2457,
+ "step": 2650
+ },
+ {
+ "epoch": 9.886005560704357,
+ "grad_norm": 149.853515625,
+ "learning_rate": 1.685254027261462e-05,
+ "loss": 6.7531,
+ "step": 2660
+ },
+ {
+ "epoch": 9.923076923076923,
+ "grad_norm": 107.76028442382812,
+ "learning_rate": 1.6790582403965305e-05,
+ "loss": 6.7149,
+ "step": 2670
+ },
+ {
+ "epoch": 9.96014828544949,
+ "grad_norm": 83.82595825195312,
+ "learning_rate": 1.6728624535315986e-05,
+ "loss": 6.4635,
+ "step": 2680
+ },
+ {
+ "epoch": 9.997219647822057,
+ "grad_norm": 106.68073272705078,
+ "learning_rate": 1.6666666666666667e-05,
+ "loss": 6.2237,
+ "step": 2690
+ },
+ {
+ "epoch": 10.037071362372567,
+ "grad_norm": 78.0780029296875,
+ "learning_rate": 1.660470879801735e-05,
+ "loss": 6.1798,
+ "step": 2700
+ },
+ {
+ "epoch": 10.037071362372567,
+ "eval_loss": 2.993870735168457,
+ "eval_runtime": 34.3527,
+ "eval_samples_per_second": 47.1,
+ "eval_steps_per_second": 6.753,
+ "step": 2700
+ },
+ {
+ "epoch": 10.074142724745135,
+ "grad_norm": 177.5696563720703,
+ "learning_rate": 1.6542750929368033e-05,
+ "loss": 7.2224,
+ "step": 2710
+ },
+ {
+ "epoch": 10.111214087117702,
+ "grad_norm": 134.7563018798828,
+ "learning_rate": 1.6480793060718714e-05,
+ "loss": 6.5327,
+ "step": 2720
+ },
+ {
+ "epoch": 10.148285449490269,
+ "grad_norm": 122.22810363769531,
+ "learning_rate": 1.6418835192069395e-05,
+ "loss": 7.4686,
+ "step": 2730
+ },
+ {
+ "epoch": 10.185356811862835,
+ "grad_norm": 137.185302734375,
+ "learning_rate": 1.6356877323420076e-05,
+ "loss": 6.1404,
+ "step": 2740
+ },
+ {
+ "epoch": 10.222428174235404,
+ "grad_norm": 100.84989166259766,
+ "learning_rate": 1.6294919454770753e-05,
+ "loss": 7.0005,
+ "step": 2750
+ },
+ {
+ "epoch": 10.25949953660797,
+ "grad_norm": 122.00653839111328,
+ "learning_rate": 1.6232961586121438e-05,
+ "loss": 5.7726,
+ "step": 2760
+ },
+ {
+ "epoch": 10.296570898980537,
+ "grad_norm": 172.2955780029297,
+ "learning_rate": 1.617100371747212e-05,
+ "loss": 6.5327,
+ "step": 2770
+ },
+ {
+ "epoch": 10.333642261353106,
+ "grad_norm": 195.36851501464844,
+ "learning_rate": 1.61090458488228e-05,
+ "loss": 7.5015,
+ "step": 2780
+ },
+ {
+ "epoch": 10.370713623725672,
+ "grad_norm": 135.3350372314453,
+ "learning_rate": 1.604708798017348e-05,
+ "loss": 6.5526,
+ "step": 2790
+ },
+ {
+ "epoch": 10.407784986098239,
+ "grad_norm": 106.01554870605469,
+ "learning_rate": 1.5985130111524162e-05,
+ "loss": 6.2078,
+ "step": 2800
+ },
+ {
+ "epoch": 10.444856348470806,
+ "grad_norm": 104.4194564819336,
+ "learning_rate": 1.5923172242874846e-05,
+ "loss": 6.1,
+ "step": 2810
+ },
+ {
+ "epoch": 10.481927710843374,
+ "grad_norm": 118.46540069580078,
+ "learning_rate": 1.5861214374225527e-05,
+ "loss": 7.1027,
+ "step": 2820
+ },
+ {
+ "epoch": 10.51899907321594,
+ "grad_norm": 108.23471069335938,
+ "learning_rate": 1.5799256505576208e-05,
+ "loss": 8.639,
+ "step": 2830
+ },
+ {
+ "epoch": 10.556070435588508,
+ "grad_norm": 144.46780395507812,
+ "learning_rate": 1.573729863692689e-05,
+ "loss": 6.9937,
+ "step": 2840
+ },
+ {
+ "epoch": 10.593141797961074,
+ "grad_norm": 108.4492416381836,
+ "learning_rate": 1.567534076827757e-05,
+ "loss": 7.2734,
+ "step": 2850
+ },
+ {
+ "epoch": 10.593141797961074,
+ "eval_loss": 2.8532466888427734,
+ "eval_runtime": 34.3677,
+ "eval_samples_per_second": 47.079,
+ "eval_steps_per_second": 6.751,
+ "step": 2850
+ },
+ {
+ "epoch": 10.630213160333643,
+ "grad_norm": 143.1586151123047,
+ "learning_rate": 1.5613382899628255e-05,
+ "loss": 7.6321,
+ "step": 2860
+ },
+ {
+ "epoch": 10.66728452270621,
+ "grad_norm": 98.63433837890625,
+ "learning_rate": 1.5551425030978936e-05,
+ "loss": 7.5788,
+ "step": 2870
+ },
+ {
+ "epoch": 10.704355885078776,
+ "grad_norm": 90.4648666381836,
+ "learning_rate": 1.5489467162329617e-05,
+ "loss": 6.7864,
+ "step": 2880
+ },
+ {
+ "epoch": 10.741427247451345,
+ "grad_norm": 141.69007873535156,
+ "learning_rate": 1.5427509293680298e-05,
+ "loss": 7.4237,
+ "step": 2890
+ },
+ {
+ "epoch": 10.778498609823911,
+ "grad_norm": 126.74280548095703,
+ "learning_rate": 1.536555142503098e-05,
+ "loss": 6.9813,
+ "step": 2900
+ },
+ {
+ "epoch": 10.815569972196478,
+ "grad_norm": 82.08953094482422,
+ "learning_rate": 1.5303593556381663e-05,
+ "loss": 6.6884,
+ "step": 2910
+ },
+ {
+ "epoch": 10.852641334569045,
+ "grad_norm": 123.8572998046875,
+ "learning_rate": 1.5241635687732344e-05,
+ "loss": 6.7464,
+ "step": 2920
+ },
+ {
+ "epoch": 10.889712696941613,
+ "grad_norm": 132.33477783203125,
+ "learning_rate": 1.5179677819083025e-05,
+ "loss": 7.7989,
+ "step": 2930
+ },
+ {
+ "epoch": 10.92678405931418,
+ "grad_norm": 81.7636489868164,
+ "learning_rate": 1.5117719950433704e-05,
+ "loss": 7.3568,
+ "step": 2940
+ },
+ {
+ "epoch": 10.963855421686747,
+ "grad_norm": 244.56277465820312,
+ "learning_rate": 1.5055762081784385e-05,
+ "loss": 8.6706,
+ "step": 2950
+ },
+ {
+ "epoch": 11.0,
+ "grad_norm": 121.42695617675781,
+ "learning_rate": 1.4993804213135068e-05,
+ "loss": 6.5687,
+ "step": 2960
+ },
+ {
+ "epoch": 11.037071362372567,
+ "grad_norm": 145.1683349609375,
+ "learning_rate": 1.493184634448575e-05,
+ "loss": 5.8992,
+ "step": 2970
+ },
+ {
+ "epoch": 11.074142724745135,
+ "grad_norm": 109.68660736083984,
+ "learning_rate": 1.4869888475836432e-05,
+ "loss": 6.4543,
+ "step": 2980
+ },
+ {
+ "epoch": 11.111214087117702,
+ "grad_norm": 93.12600708007812,
+ "learning_rate": 1.4807930607187113e-05,
+ "loss": 6.1386,
+ "step": 2990
+ },
+ {
+ "epoch": 11.148285449490269,
+ "grad_norm": 167.23182678222656,
+ "learning_rate": 1.4745972738537794e-05,
+ "loss": 6.9047,
+ "step": 3000
+ },
+ {
+ "epoch": 11.148285449490269,
+ "eval_loss": 2.9146969318389893,
+ "eval_runtime": 34.386,
+ "eval_samples_per_second": 47.054,
+ "eval_steps_per_second": 6.747,
+ "step": 3000
+ }
+ ],
+ "logging_steps": 10,
+ "max_steps": 5380,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 20,
+ "save_steps": 150,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 0.0,
+ "train_batch_size": 7,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-3000/training_args.bin b/checkpoints/checkpoint-3000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..f1a69c7344d84f80932397ffdc7afef25687d5b4
--- /dev/null
+++ b/checkpoints/checkpoint-3000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fda94467724e48ac975735648aa63a75558d48b6c993506e918a3476acf80697
+size 129
diff --git a/checkpoints/checkpoint-4050/1_Pooling/config.json b/checkpoints/checkpoint-4050/1_Pooling/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b68441c3f37d8bee79501ca0afe536bf19753928
--- /dev/null
+++ b/checkpoints/checkpoint-4050/1_Pooling/config.json
@@ -0,0 +1,10 @@
+{
+ "word_embedding_dimension": 1024,
+ "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
+}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-4050/README.md b/checkpoints/checkpoint-4050/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..4c4446adf7d585427e1d1e9a7ec7731373d7a18b
--- /dev/null
+++ b/checkpoints/checkpoint-4050/README.md
@@ -0,0 +1,962 @@
+---
+tags:
+- sentence-transformers
+- sentence-similarity
+- feature-extraction
+- generated_from_trainer
+- dataset_size:7552
+- loss:CoSENTLoss
+base_model: intfloat/multilingual-e5-large-instruct
+widget:
+- source_sentence: How are calibration points linked to equipment?
+ sentences:
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+ - "How does a flow computer generate and store reports?\nA flow computer generates\
+ \ daily or hourly reports to provide users with operational data. These reports\
+ \ are stored in the flow computer's memory in an organized format.\n\nReport structure:\n\
+ - Each report includes:\n- Date and time of the data recording.\n- Data recorded\
+ \ from flow computers.\n\nData storage in tables:\nThe reports are saved in two\
+ \ tables:\n1. Main table (Index):\n - Stores the date, time, and flow computer\
+ \ identifier.\n2. Detail table:\n - Stores the measured values associated with\
+ \ the report.\n\nConnection to the Modbus table:\nThe flow computer's reports\
+ \ are linked to a Modbus table. This table contains the names corresponding to\
+ \ each value in the reports, making it easier to interpret the data."
+ - "What is uncertainty?\nUncertainty is a measure of confidence in the precision\
+ \ and reliability of results obtained from equipment or measurement systems. It\
+ \ quantifies the potential error or margin of error in measurements.\n\nTypes\
+ \ of uncertainty:\nThere are two main types of uncertainty:\n1. Uncertainty of\
+ \ magnitudes (variables):\n - Refers to the uncertainty of specific variables,\
+ \ such as temperature or pressure.\n - It is calculated after calibrating a\
+ \ device or obtained from the equipment manufacturer's manual.\n - This uncertainty\
+ \ serves as a starting point for further calculations related to the equipment.\n\
+ \n2. Uncertainty of the measurement system:\n - Refers to the uncertainty calculated\
+ \ for the overall flow measurement.\n - It depends on the uncertainties of\
+ \ the individual variables (magnitudes) and represents the combined margin of\
+ \ error for the entire system.\n\nKey points:\n- The uncertainties of magnitudes\
+ \ (variables) are the foundation for calculating the uncertainty of the measurement\
+ \ system. Think of them as the \"building blocks.\"\n- Do not confuse the two\
+ \ types of uncertainty:\n - **Uncertainty of magnitudes/variables**: Specific\
+ \ to individual variables (e.g., temperature, pressure).\n - **Uncertainty\
+ \ of the measurement system**: Specific to the overall flow measurement.\n\nDatabase\
+ \ storage for uncertainties:\nIn the database, uncertainty calculations are stored\
+ \ in two separate tables:\n1. Uncertainty of magnitudes (variables):\n - Stores\
+ \ the uncertainty values for specific variables (e.g., temperature, pressure).\n\
+ \n2. Uncertainty of the measurement system:\n - Stores the uncertainty values\
+ \ for the overall flow measurement system.\n\nHow to retrieve uncertainty data:\n\
+ - To find the uncertainty of the measurement system, join the measurement systems\
+ \ table with the uncertainty of the measurement system table.\n- To find the uncertainty\
+ \ of a specific variable (magnitude), join the measurement systems table with\
+ \ the uncertainty of magnitudes (variables) table.\n\nImportant note:\nDo not\
+ \ confuse the two types of uncertainty:\n- If the user requests the uncertainty\
+ \ of the measurement system, use the first join (measurement systems table + uncertainty\
+ \ of the measurement system table).\n- If the user requests the uncertainty of\
+ \ a specific variable (magnitude) in a report, use the second join (measurement\
+ \ systems table + uncertainty of magnitudes table)."
+- source_sentence: What is the primary key of the flow computer table?
+ sentences:
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - "How does a flow computer generate and store reports?\nA flow computer generates\
+ \ daily or hourly reports to provide users with operational data. These reports\
+ \ are stored in the flow computer's memory in an organized format.\n\nReport structure:\n\
+ - Each report includes:\n- Date and time of the data recording.\n- Data recorded\
+ \ from flow computers.\n\nData storage in tables:\nThe reports are saved in two\
+ \ tables:\n1. Main table (Index):\n - Stores the date, time, and flow computer\
+ \ identifier.\n2. Detail table:\n - Stores the measured values associated with\
+ \ the report.\n\nConnection to the Modbus table:\nThe flow computer's reports\
+ \ are linked to a Modbus table. This table contains the names corresponding to\
+ \ each value in the reports, making it easier to interpret the data."
+- source_sentence: Can you provide a sample query to test the retrieval of the uncertainty
+ result for the specified tag and date?
+ sentences:
+ - 'What is equipment calibration?
+
+ Calibration is a metrological verification process used to ensure the accuracy
+ of measurement equipment. It is performed periodically, based on intervals set
+ by the company or a regulatory body.
+
+
+ Purpose of calibration:
+
+ The calibration process corrects any deviations in how the equipment measures
+ physical magnitudes (variables). This ensures the equipment provides accurate
+ and reliable data.
+
+
+ Calibration cycles:
+
+ There are two main calibration cycles:
+
+ 1. As-found: Represents the equipment''s measurement accuracy before any adjustments
+ are made. This cycle is almost always implemented.
+
+ 2. As-left: Represents the equipment''s measurement accuracy after adjustments
+ are made. This cycle is used depending on regulatory requirements.
+
+
+ Calibration uncertainty:
+
+ - Uncertainty is included in the results of a calibration.
+
+ - Calibration uncertainty refers to the margin of error in the device''s measurements,
+ which also affects the uncertainty of the measured variable or magnitude.'
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - "What is uncertainty?\nUncertainty is a measure of confidence in the precision\
+ \ and reliability of results obtained from equipment or measurement systems. It\
+ \ quantifies the potential error or margin of error in measurements.\n\nTypes\
+ \ of uncertainty:\nThere are two main types of uncertainty:\n1. Uncertainty of\
+ \ magnitudes (variables):\n - Refers to the uncertainty of specific variables,\
+ \ such as temperature or pressure.\n - It is calculated after calibrating a\
+ \ device or obtained from the equipment manufacturer's manual.\n - This uncertainty\
+ \ serves as a starting point for further calculations related to the equipment.\n\
+ \n2. Uncertainty of the measurement system:\n - Refers to the uncertainty calculated\
+ \ for the overall flow measurement.\n - It depends on the uncertainties of\
+ \ the individual variables (magnitudes) and represents the combined margin of\
+ \ error for the entire system.\n\nKey points:\n- The uncertainties of magnitudes\
+ \ (variables) are the foundation for calculating the uncertainty of the measurement\
+ \ system. Think of them as the \"building blocks.\"\n- Do not confuse the two\
+ \ types of uncertainty:\n - **Uncertainty of magnitudes/variables**: Specific\
+ \ to individual variables (e.g., temperature, pressure).\n - **Uncertainty\
+ \ of the measurement system**: Specific to the overall flow measurement.\n\nDatabase\
+ \ storage for uncertainties:\nIn the database, uncertainty calculations are stored\
+ \ in two separate tables:\n1. Uncertainty of magnitudes (variables):\n - Stores\
+ \ the uncertainty values for specific variables (e.g., temperature, pressure).\n\
+ \n2. Uncertainty of the measurement system:\n - Stores the uncertainty values\
+ \ for the overall flow measurement system.\n\nHow to retrieve uncertainty data:\n\
+ - To find the uncertainty of the measurement system, join the measurement systems\
+ \ table with the uncertainty of the measurement system table.\n- To find the uncertainty\
+ \ of a specific variable (magnitude), join the measurement systems table with\
+ \ the uncertainty of magnitudes (variables) table.\n\nImportant note:\nDo not\
+ \ confuse the two types of uncertainty:\n- If the user requests the uncertainty\
+ \ of the measurement system, use the first join (measurement systems table + uncertainty\
+ \ of the measurement system table).\n- If the user requests the uncertainty of\
+ \ a specific variable (magnitude) in a report, use the second join (measurement\
+ \ systems table + uncertainty of magnitudes table)."
+- source_sentence: How are the secondary equipment and measurement system related?
+ sentences:
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - 'What do measurement equipment measure?
+
+ Each equipment measures a physical magnitude, also known as a variable. Based
+ on the type of variable they measure, devices are classified into different categories.
+
+
+ Equipment classification:
+
+ - Primary meter: Assigned by default to equipments like orifice plates.
+
+ - Secondary meter: Assigned by default to equipments like transmitters.
+
+ - Tertiary meter: Used for other types of equipments.
+
+
+ Equipment types in the database:
+
+ The database includes a table listing all equipment types. Examples of equipment
+ types are:
+
+ - Differential pressure transmitters
+
+ - RTDs (Resistance Temperature Detectors)
+
+ - Orifice plates
+
+ - Multivariable transmitters
+
+ - Ultrasonic meters
+
+
+ Meteorological checks for equipments:
+
+ Each equipment type is assigned a meteorological check, which can be either:
+
+ - Calibration: To ensure measurement accuracy.
+
+ - Inspection: To verify proper functioning.
+
+
+ Data storage in tables:
+
+ The database also includes a separate table for equipment classifications, which
+ are:
+
+ - Primary meter
+
+ - Secondary meter
+
+ - Tertiary meter
+
+ So, an equipment has equipment types and this types has classifications.'
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+- source_sentence: What is the table structure for secondary equipment?
+ sentences:
+ - 'What kind of data store an equipment?
+
+ Equipments can capture meteorological data, such as pressure, temperature, and
+ volume (magnitudes). This data is essential for users to perform various calculations.
+
+
+ Data storage:
+
+ - The measured values are stored in a special table in the database for magnitudes.
+ This table contains the values of the variables captured by the equipments.
+
+ - These values are **direct measurements** from the fluid (e.g., raw pressure,
+ temperature, or volume readings). **They are not calculated values**, such as
+ uncertainty.
+
+ - The values stored in the variable values table are **different** from variable
+ uncertainty values, which are calculated separately and represent the margin of
+ error.
+
+
+ Accessing the data:
+
+ - Users typically access the data by referring to the readings from the measurement
+ system, not directly from the individual equipments.
+
+ - The readings are stored in a "variable values" table within the database.
+
+
+ Linking variable names:
+
+ If the user needs to know the name of a variable, they must link the data to another
+ table that stores information about the types of variables.'
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+ - 'How are flow computers and measurement systems related?
+
+ Flow computers can have multiple systems assigned to them. However, a measurement
+ system can only be assigned to one flow computer.
+
+
+ Database terminology:
+
+ In the database, this relationship is referred to as:
+
+ - Meter streams
+
+ - Meter runs
+
+ - Sections
+
+
+ Storage of the relationship:
+
+ The relationship between a flow computer and its assigned measurement system is
+ stored in a special table.
+
+
+ User context:
+
+ When a user refers to a "meter stream," they are indicating that they are searching
+ for a measurement system assigned to a specific flow computer.'
+datasets:
+- Lauther/measuring-embeddings-v3
+pipeline_tag: sentence-similarity
+library_name: sentence-transformers
+---
+
+# SentenceTransformer based on intfloat/multilingual-e5-large-instruct
+
+This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [intfloat/multilingual-e5-large-instruct](https://huggingface.co/intfloat/multilingual-e5-large-instruct) on the [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) 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.
+
+## Model Details
+
+### Model Description
+- **Model Type:** Sentence Transformer
+- **Base model:** [intfloat/multilingual-e5-large-instruct](https://huggingface.co/intfloat/multilingual-e5-large-instruct)
+- **Maximum Sequence Length:** 512 tokens
+- **Output Dimensionality:** 1024 dimensions
+- **Similarity Function:** Cosine Similarity
+- **Training Dataset:**
+ - [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3)
+
+
+
+### Model Sources
+
+- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
+- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
+- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
+
+### Full Model Architecture
+
+```
+SentenceTransformer(
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: XLMRobertaModel
+ (1): Pooling({'word_embedding_dimension': 1024, '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()
+)
+```
+
+## Usage
+
+### Direct Usage (Sentence Transformers)
+
+First install the Sentence Transformers library:
+
+```bash
+pip install -U sentence-transformers
+```
+
+Then you can load this model and run inference.
+```python
+from sentence_transformers import SentenceTransformer
+
+# Download from the 🤗 Hub
+model = SentenceTransformer("sentence_transformers_model_id")
+# Run inference
+sentences = [
+ 'What is the table structure for secondary equipment?',
+ 'How are flow computers and measurement systems related?\nFlow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.\n\nDatabase terminology:\nIn the database, this relationship is referred to as:\n- Meter streams\n- Meter runs\n- Sections\n\nStorage of the relationship:\nThe relationship between a flow computer and its assigned measurement system is stored in a special table.\n\nUser context:\nWhen a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer.',
+ 'What kind of data store an equipment?\nEquipments can capture meteorological data, such as pressure, temperature, and volume (magnitudes). This data is essential for users to perform various calculations.\n\nData storage:\n- The measured values are stored in a special table in the database for magnitudes. This table contains the values of the variables captured by the equipments.\n- These values are **direct measurements** from the fluid (e.g., raw pressure, temperature, or volume readings). **They are not calculated values**, such as uncertainty.\n- The values stored in the variable values table are **different** from variable uncertainty values, which are calculated separately and represent the margin of error.\n\nAccessing the data:\n- Users typically access the data by referring to the readings from the measurement system, not directly from the individual equipments.\n- The readings are stored in a "variable values" table within the database.\n\nLinking variable names:\nIf the user needs to know the name of a variable, they must link the data to another table that stores information about the types of variables.',
+]
+embeddings = model.encode(sentences)
+print(embeddings.shape)
+# [3, 1024]
+
+# Get the similarity scores for the embeddings
+similarities = model.similarity(embeddings, embeddings)
+print(similarities.shape)
+# [3, 3]
+```
+
+
+
+
+
+
+
+
+
+
+
+## Training Details
+
+### Training Dataset
+
+#### measuring-embeddings-v3
+
+* Dataset: [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) at [1b3cbbe](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3/tree/1b3cbbeb70b63338110491cd3de2950fb40b4f87)
+* Size: 7,552 training samples
+* Columns: sentence1, sentence2, and score
+* Approximate statistics based on the first 1000 samples:
+ | | sentence1 | sentence2 | score |
+ |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:----------------------------------------------------------------|
+ | type | string | string | float |
+ | details | - min: 9 tokens
- mean: 15.96 tokens
- max: 40 tokens
| - min: 120 tokens
- mean: 255.56 tokens
- max: 512 tokens
| - min: 0.0
- mean: 0.22
- max: 0.95
|
+* Samples:
+ | sentence1 | sentence2 | score |
+ |:-------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------|
+ | How can I combine the sub-query with the main query to fetch the last uncertainty report? | What do measurement equipment measure?
Each equipment measures a physical magnitude, also known as a variable. Based on the type of variable they measure, devices are classified into different categories.
Equipment classification:
- Primary meter: Assigned by default to equipments like orifice plates.
- Secondary meter: Assigned by default to equipments like transmitters.
- Tertiary meter: Used for other types of equipments.
Equipment types in the database:
The database includes a table listing all equipment types. Examples of equipment types are:
- Differential pressure transmitters
- RTDs (Resistance Temperature Detectors)
- Orifice plates
- Multivariable transmitters
- Ultrasonic meters
Meteorological checks for equipments:
Each equipment type is assigned a meteorological check, which can be either:
- Calibration: To ensure measurement accuracy.
- Inspection: To verify proper functioning.
Data storage in tables:
The database also includes a separate table for equipment classific... | 0.1 |
+ | What is the column name for the calibration date in the calibration table? | How are flow computers and measurement systems related?
Flow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.
Database terminology:
In the database, this relationship is referred to as:
- Meter streams
- Meter runs
- Sections
Storage of the relationship:
The relationship between a flow computer and its assigned measurement system is stored in a special table.
User context:
When a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer. | 0.1 |
+ | What is the name of the table that contains the flow computer tags? | What is equipment calibration?
Calibration is a metrological verification process used to ensure the accuracy of measurement equipment. It is performed periodically, based on intervals set by the company or a regulatory body.
Purpose of calibration:
The calibration process corrects any deviations in how the equipment measures physical magnitudes (variables). This ensures the equipment provides accurate and reliable data.
Calibration cycles:
There are two main calibration cycles:
1. As-found: Represents the equipment's measurement accuracy before any adjustments are made. This cycle is almost always implemented.
2. As-left: Represents the equipment's measurement accuracy after adjustments are made. This cycle is used depending on regulatory requirements.
Calibration uncertainty:
- Uncertainty is included in the results of a calibration.
- Calibration uncertainty refers to the margin of error in the device's measurements, which also affects the uncertainty of the measured variable or ... | 0.05 |
+* Loss: [CoSENTLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
+ ```json
+ {
+ "scale": 20.0,
+ "similarity_fct": "pairwise_cos_sim"
+ }
+ ```
+
+### Evaluation Dataset
+
+#### measuring-embeddings-v3
+
+* Dataset: [measuring-embeddings-v3](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3) at [1b3cbbe](https://huggingface.co/datasets/Lauther/measuring-embeddings-v3/tree/1b3cbbeb70b63338110491cd3de2950fb40b4f87)
+* Size: 1,618 evaluation samples
+* Columns: sentence1, sentence2, and score
+* Approximate statistics based on the first 1000 samples:
+ | | sentence1 | sentence2 | score |
+ |:--------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:----------------------------------------------------------------|
+ | type | string | string | float |
+ | details | - min: 9 tokens
- mean: 15.83 tokens
- max: 40 tokens
| - min: 120 tokens
- mean: 250.41 tokens
- max: 512 tokens
| - min: 0.0
- mean: 0.23
- max: 0.95
|
+* Samples:
+ | sentence1 | sentence2 | score |
+ |:--------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------|
+ | Identify any additional tables or columns that might be needed for the query. | How are flow computers and measurement systems related?
Flow computers can have multiple systems assigned to them. However, a measurement system can only be assigned to one flow computer.
Database terminology:
In the database, this relationship is referred to as:
- Meter streams
- Meter runs
- Sections
Storage of the relationship:
The relationship between a flow computer and its assigned measurement system is stored in a special table.
User context:
When a user refers to a "meter stream," they are indicating that they are searching for a measurement system assigned to a specific flow computer. | 0.2 |
+ | What columns in these tables contain the measurement system tag and the flow computer tag? | How does a flow computer generate and store reports?
A flow computer generates daily or hourly reports to provide users with operational data. These reports are stored in the flow computer's memory in an organized format.
Report structure:
- Each report includes:
- Date and time of the data recording.
- Data recorded from flow computers.
Data storage in tables:
The reports are saved in two tables:
1. Main table (Index):
- Stores the date, time, and flow computer identifier.
2. Detail table:
- Stores the measured values associated with the report.
Connection to the Modbus table:
The flow computer's reports are linked to a Modbus table. This table contains the names corresponding to each value in the reports, making it easier to interpret the data. | 0.1 |
+ | Identify the column that stores the calibration number. | What kind of data store an equipment?
Equipments can capture meteorological data, such as pressure, temperature, and volume (magnitudes). This data is essential for users to perform various calculations.
Data storage:
- The measured values are stored in a special table in the database for magnitudes. This table contains the values of the variables captured by the equipments.
- These values are **direct measurements** from the fluid (e.g., raw pressure, temperature, or volume readings). **They are not calculated values**, such as uncertainty.
- The values stored in the variable values table are **different** from variable uncertainty values, which are calculated separately and represent the margin of error.
Accessing the data:
- Users typically access the data by referring to the readings from the measurement system, not directly from the individual equipments.
- The readings are stored in a "variable values" table within the database.
Linking variable names:
If the user needs to kno... | 0.1 |
+* Loss: [CoSENTLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
+ ```json
+ {
+ "scale": 20.0,
+ "similarity_fct": "pairwise_cos_sim"
+ }
+ ```
+
+### Training Hyperparameters
+#### Non-Default Hyperparameters
+
+- `eval_strategy`: steps
+- `per_device_train_batch_size`: 7
+- `per_device_eval_batch_size`: 7
+- `gradient_accumulation_steps`: 4
+- `learning_rate`: 3e-05
+- `num_train_epochs`: 20
+- `warmup_ratio`: 0.1
+
+#### All Hyperparameters
+Click to expand
+
+- `overwrite_output_dir`: False
+- `do_predict`: False
+- `eval_strategy`: steps
+- `prediction_loss_only`: True
+- `per_device_train_batch_size`: 7
+- `per_device_eval_batch_size`: 7
+- `per_gpu_train_batch_size`: None
+- `per_gpu_eval_batch_size`: None
+- `gradient_accumulation_steps`: 4
+- `eval_accumulation_steps`: None
+- `torch_empty_cache_steps`: None
+- `learning_rate`: 3e-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`: 20
+- `max_steps`: -1
+- `lr_scheduler_type`: linear
+- `lr_scheduler_kwargs`: {}
+- `warmup_ratio`: 0.1
+- `warmup_steps`: 0
+- `log_level`: passive
+- `log_level_replica`: warning
+- `log_on_each_node`: True
+- `logging_nan_inf_filter`: True
+- `save_safetensors`: True
+- `save_on_each_node`: False
+- `save_only_model`: False
+- `restore_callback_states_from_checkpoint`: False
+- `no_cuda`: False
+- `use_cpu`: False
+- `use_mps_device`: False
+- `seed`: 42
+- `data_seed`: None
+- `jit_mode_eval`: False
+- `use_ipex`: False
+- `bf16`: False
+- `fp16`: False
+- `fp16_opt_level`: O1
+- `half_precision_backend`: auto
+- `bf16_full_eval`: False
+- `fp16_full_eval`: False
+- `tf32`: None
+- `local_rank`: 0
+- `ddp_backend`: None
+- `tpu_num_cores`: None
+- `tpu_metrics_debug`: False
+- `debug`: []
+- `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`: False
+- `ignore_data_skip`: False
+- `fsdp`: []
+- `fsdp_min_num_params`: 0
+- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
+- `fsdp_transformer_layer_cls_to_wrap`: None
+- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
+- `deepspeed`: None
+- `label_smoothing_factor`: 0.0
+- `optim`: adamw_torch
+- `optim_args`: None
+- `adafactor`: False
+- `group_by_length`: False
+- `length_column_name`: length
+- `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`: False
+- `resume_from_checkpoint`: None
+- `hub_model_id`: None
+- `hub_strategy`: every_save
+- `hub_private_repo`: None
+- `hub_always_push`: False
+- `gradient_checkpointing`: False
+- `gradient_checkpointing_kwargs`: None
+- `include_inputs_for_metrics`: False
+- `include_for_metrics`: []
+- `eval_do_concat_batches`: True
+- `fp16_backend`: auto
+- `push_to_hub_model_id`: None
+- `push_to_hub_organization`: None
+- `mp_parameters`:
+- `auto_find_batch_size`: False
+- `full_determinism`: False
+- `torchdynamo`: None
+- `ray_scope`: last
+- `ddp_timeout`: 1800
+- `torch_compile`: False
+- `torch_compile_backend`: None
+- `torch_compile_mode`: None
+- `dispatch_batches`: None
+- `split_batches`: None
+- `include_tokens_per_second`: False
+- `include_num_input_tokens_seen`: False
+- `neftune_noise_alpha`: None
+- `optim_target_modules`: None
+- `batch_eval_metrics`: False
+- `eval_on_start`: False
+- `use_liger_kernel`: False
+- `eval_use_gather_object`: False
+- `average_tokens_across_devices`: False
+- `prompts`: None
+- `batch_sampler`: batch_sampler
+- `multi_dataset_batch_sampler`: proportional
+
+Click to expand
+
+| Epoch | Step | Training Loss | Validation Loss |
+|:-------:|:----:|:-------------:|:---------------:|
+| 9.5153 | 2560 | 6.782 | - |
+| 9.5524 | 2570 | 7.3027 | - |
+| 9.5894 | 2580 | 7.3348 | - |
+| 9.6265 | 2590 | 7.7864 | - |
+| 9.6636 | 2600 | 6.3552 | - |
+| 9.7006 | 2610 | 7.151 | - |
+| 9.7377 | 2620 | 6.1664 | - |
+| 9.7748 | 2630 | 6.0398 | - |
+| 9.8119 | 2640 | 7.0452 | - |
+| 9.8489 | 2650 | 7.2457 | - |
+| 9.8860 | 2660 | 6.7531 | - |
+| 9.9231 | 2670 | 6.7149 | - |
+| 9.9601 | 2680 | 6.4635 | - |
+| 9.9972 | 2690 | 6.2237 | - |
+| 10.0371 | 2700 | 6.1798 | 2.9939 |
+| 10.0741 | 2710 | 7.2224 | - |
+| 10.1112 | 2720 | 6.5327 | - |
+| 10.1483 | 2730 | 7.4686 | - |
+| 10.1854 | 2740 | 6.1404 | - |
+| 10.2224 | 2750 | 7.0005 | - |
+| 10.2595 | 2760 | 5.7726 | - |
+| 10.2966 | 2770 | 6.5327 | - |
+| 10.3336 | 2780 | 7.5015 | - |
+| 10.3707 | 2790 | 6.5526 | - |
+| 10.4078 | 2800 | 6.2078 | - |
+| 10.4449 | 2810 | 6.1 | - |
+| 10.4819 | 2820 | 7.1027 | - |
+| 10.5190 | 2830 | 8.639 | - |
+| 10.5561 | 2840 | 6.9937 | - |
+| 10.5931 | 2850 | 7.2734 | 2.8532 |
+| 10.6302 | 2860 | 7.6321 | - |
+| 10.6673 | 2870 | 7.5788 | - |
+| 10.7044 | 2880 | 6.7864 | - |
+| 10.7414 | 2890 | 7.4237 | - |
+| 10.7785 | 2900 | 6.9813 | - |
+| 10.8156 | 2910 | 6.6884 | - |
+| 10.8526 | 2920 | 6.7464 | - |
+| 10.8897 | 2930 | 7.7989 | - |
+| 10.9268 | 2940 | 7.3568 | - |
+| 10.9639 | 2950 | 8.6706 | - |
+| 11.0 | 2960 | 6.5687 | - |
+| 11.0371 | 2970 | 5.8992 | - |
+| 11.0741 | 2980 | 6.4543 | - |
+| 11.1112 | 2990 | 6.1386 | - |
+| 11.1483 | 3000 | 6.9047 | 2.9147 |
+| 11.1854 | 3010 | 7.405 | - |
+| 11.2224 | 3020 | 7.5441 | - |
+| 11.2595 | 3030 | 6.7524 | - |
+| 11.2966 | 3040 | 7.698 | - |
+| 11.3336 | 3050 | 7.6167 | - |
+| 11.3707 | 3060 | 7.1516 | - |
+| 11.4078 | 3070 | 6.7458 | - |
+| 11.4449 | 3080 | 6.7608 | - |
+| 11.4819 | 3090 | 7.1508 | - |
+| 11.5190 | 3100 | 6.9155 | - |
+| 11.5561 | 3110 | 6.6664 | - |
+| 11.5931 | 3120 | 8.3841 | - |
+| 11.6302 | 3130 | 7.1934 | - |
+| 11.6673 | 3140 | 6.9681 | - |
+| 11.7044 | 3150 | 7.2187 | 2.7509 |
+| 11.7414 | 3160 | 7.3155 | - |
+| 11.7785 | 3170 | 7.3103 | - |
+| 11.8156 | 3180 | 7.1959 | - |
+| 11.8526 | 3190 | 6.8164 | - |
+| 11.8897 | 3200 | 7.5836 | - |
+| 11.9268 | 3210 | 5.2671 | - |
+| 11.9639 | 3220 | 6.4929 | - |
+| 12.0 | 3230 | 7.0892 | - |
+| 12.0371 | 3240 | 7.0877 | - |
+| 12.0741 | 3250 | 5.8302 | - |
+| 12.1112 | 3260 | 5.6145 | - |
+| 12.1483 | 3270 | 6.5808 | - |
+| 12.1854 | 3280 | 6.6826 | - |
+| 12.2224 | 3290 | 5.9819 | - |
+| 12.2595 | 3300 | 6.68 | 3.0175 |
+| 12.2966 | 3310 | 6.1685 | - |
+| 12.3336 | 3320 | 6.4473 | - |
+| 12.3707 | 3330 | 6.3965 | - |
+| 12.4078 | 3340 | 6.6278 | - |
+| 12.4449 | 3350 | 5.4575 | - |
+| 12.4819 | 3360 | 7.3019 | - |
+| 12.5190 | 3370 | 7.4843 | - |
+| 12.5561 | 3380 | 6.709 | - |
+| 12.5931 | 3390 | 6.7168 | - |
+| 12.6302 | 3400 | 7.0223 | - |
+| 12.6673 | 3410 | 6.5089 | - |
+| 12.7044 | 3420 | 6.5094 | - |
+| 12.7414 | 3430 | 7.2317 | - |
+| 12.7785 | 3440 | 6.6885 | - |
+| 12.8156 | 3450 | 6.9693 | 2.8462 |
+| 12.8526 | 3460 | 6.8242 | - |
+| 12.8897 | 3470 | 6.6899 | - |
+| 12.9268 | 3480 | 6.9113 | - |
+| 12.9639 | 3490 | 7.1903 | - |
+| 13.0 | 3500 | 7.3286 | - |
+| 13.0371 | 3510 | 6.5465 | - |
+| 13.0741 | 3520 | 5.6804 | - |
+| 13.1112 | 3530 | 5.6412 | - |
+| 13.1483 | 3540 | 6.6161 | - |
+| 13.1854 | 3550 | 5.761 | - |
+| 13.2224 | 3560 | 5.5669 | - |
+| 13.2595 | 3570 | 5.6184 | - |
+| 13.2966 | 3580 | 6.2996 | - |
+| 13.3336 | 3590 | 4.99 | - |
+| 13.3707 | 3600 | 5.9974 | 3.2358 |
+| 13.4078 | 3610 | 5.6962 | - |
+| 13.4449 | 3620 | 6.3662 | - |
+| 13.4819 | 3630 | 7.0398 | - |
+| 13.5190 | 3640 | 7.7358 | - |
+| 13.5561 | 3650 | 7.9063 | - |
+| 13.5931 | 3660 | 5.7823 | - |
+| 13.6302 | 3670 | 6.9861 | - |
+| 13.6673 | 3680 | 7.2855 | - |
+| 13.7044 | 3690 | 5.6785 | - |
+| 13.7414 | 3700 | 6.4071 | - |
+| 13.7785 | 3710 | 6.4294 | - |
+| 13.8156 | 3720 | 6.0842 | - |
+| 13.8526 | 3730 | 5.9422 | - |
+| 13.8897 | 3740 | 7.0778 | - |
+| 13.9268 | 3750 | 8.1597 | 3.0093 |
+| 13.9639 | 3760 | 6.3154 | - |
+| 14.0 | 3770 | 6.2416 | - |
+| 14.0371 | 3780 | 5.9958 | - |
+| 14.0741 | 3790 | 5.7032 | - |
+| 14.1112 | 3800 | 4.9524 | - |
+| 14.1483 | 3810 | 5.386 | - |
+| 14.1854 | 3820 | 5.6353 | - |
+| 14.2224 | 3830 | 5.0873 | - |
+| 14.2595 | 3840 | 4.9255 | - |
+| 14.2966 | 3850 | 5.1423 | - |
+| 14.3336 | 3860 | 6.0775 | - |
+| 14.3707 | 3870 | 4.5073 | - |
+| 14.4078 | 3880 | 6.8347 | - |
+| 14.4449 | 3890 | 6.5397 | - |
+| 14.4819 | 3900 | 7.2143 | 3.3080 |
+| 14.5190 | 3910 | 6.1123 | - |
+| 14.5561 | 3920 | 6.6048 | - |
+| 14.5931 | 3930 | 6.3464 | - |
+| 14.6302 | 3940 | 6.3618 | - |
+| 14.6673 | 3950 | 6.5718 | - |
+| 14.7044 | 3960 | 5.9785 | - |
+| 14.7414 | 3970 | 6.5758 | - |
+| 14.7785 | 3980 | 6.4308 | - |
+| 14.8156 | 3990 | 6.0208 | - |
+| 14.8526 | 4000 | 6.0303 | - |
+| 14.8897 | 4010 | 6.6396 | - |
+| 14.9268 | 4020 | 6.0184 | - |
+| 14.9639 | 4030 | 6.6248 | - |
+| 15.0 | 4040 | 6.4538 | - |
+| 15.0371 | 4050 | 6.4742 | 3.1761 |
+
+",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "cls_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "eos_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "mask_token": {
+ "content": "",
+ "lstrip": true,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "pad_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "sep_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ },
+ "unk_token": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false
+ }
+}
diff --git a/checkpoints/checkpoint-4050/tokenizer.json b/checkpoints/checkpoint-4050/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ed19c669b24d8e44e6e4c98b8ef4bb83d8cab46
--- /dev/null
+++ b/checkpoints/checkpoint-4050/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de6f09c3f9b891e5b98dd3af9463dcab5a97d5265e288271395324a0577e6c05
+size 133
diff --git a/checkpoints/checkpoint-4050/tokenizer_config.json b/checkpoints/checkpoint-4050/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0df9458a41642f96287ae84a989281a3bdcebacf
--- /dev/null
+++ b/checkpoints/checkpoint-4050/tokenizer_config.json
@@ -0,0 +1,56 @@
+{
+ "added_tokens_decoder": {
+ "0": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "1": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "2": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "3": {
+ "content": "",
+ "lstrip": false,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ },
+ "250001": {
+ "content": "",
+ "lstrip": true,
+ "normalized": false,
+ "rstrip": false,
+ "single_word": false,
+ "special": true
+ }
+ },
+ "additional_special_tokens": [],
+ "bos_token": "",
+ "clean_up_tokenization_spaces": true,
+ "cls_token": "",
+ "eos_token": "",
+ "extra_special_tokens": {},
+ "mask_token": "",
+ "model_max_length": 512,
+ "pad_token": "",
+ "sep_token": "",
+ "tokenizer_class": "XLMRobertaTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoints/checkpoint-4050/trainer_state.json b/checkpoints/checkpoint-4050/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..7eefbc154233415c3bcd47ae361cac154f00e4e7
--- /dev/null
+++ b/checkpoints/checkpoint-4050/trainer_state.json
@@ -0,0 +1,3084 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 15.037071362372567,
+ "eval_steps": 150,
+ "global_step": 4050,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.03707136237256719,
+ "grad_norm": 53.36411666870117,
+ "learning_rate": 1.1152416356877324e-06,
+ "loss": 10.8165,
+ "step": 10
+ },
+ {
+ "epoch": 0.07414272474513438,
+ "grad_norm": 61.33858871459961,
+ "learning_rate": 2.2304832713754648e-06,
+ "loss": 10.8463,
+ "step": 20
+ },
+ {
+ "epoch": 0.11121408711770157,
+ "grad_norm": 67.16305541992188,
+ "learning_rate": 3.345724907063197e-06,
+ "loss": 10.821,
+ "step": 30
+ },
+ {
+ "epoch": 0.14828544949026876,
+ "grad_norm": 62.08958053588867,
+ "learning_rate": 4.4609665427509296e-06,
+ "loss": 10.4704,
+ "step": 40
+ },
+ {
+ "epoch": 0.18535681186283595,
+ "grad_norm": 53.839927673339844,
+ "learning_rate": 5.576208178438661e-06,
+ "loss": 10.8568,
+ "step": 50
+ },
+ {
+ "epoch": 0.22242817423540315,
+ "grad_norm": 57.02272033691406,
+ "learning_rate": 6.691449814126394e-06,
+ "loss": 10.3005,
+ "step": 60
+ },
+ {
+ "epoch": 0.2594995366079703,
+ "grad_norm": 88.16449737548828,
+ "learning_rate": 7.806691449814127e-06,
+ "loss": 9.6009,
+ "step": 70
+ },
+ {
+ "epoch": 0.2965708989805375,
+ "grad_norm": 84.35252380371094,
+ "learning_rate": 8.921933085501859e-06,
+ "loss": 9.7448,
+ "step": 80
+ },
+ {
+ "epoch": 0.3336422613531047,
+ "grad_norm": 77.93982696533203,
+ "learning_rate": 1.0037174721189593e-05,
+ "loss": 9.5922,
+ "step": 90
+ },
+ {
+ "epoch": 0.3707136237256719,
+ "grad_norm": 56.669166564941406,
+ "learning_rate": 1.1152416356877323e-05,
+ "loss": 10.0919,
+ "step": 100
+ },
+ {
+ "epoch": 0.4077849860982391,
+ "grad_norm": 56.652488708496094,
+ "learning_rate": 1.2267657992565056e-05,
+ "loss": 10.2733,
+ "step": 110
+ },
+ {
+ "epoch": 0.4448563484708063,
+ "grad_norm": 63.472042083740234,
+ "learning_rate": 1.3382899628252788e-05,
+ "loss": 9.4547,
+ "step": 120
+ },
+ {
+ "epoch": 0.4819277108433735,
+ "grad_norm": 78.0159912109375,
+ "learning_rate": 1.4498141263940521e-05,
+ "loss": 9.1738,
+ "step": 130
+ },
+ {
+ "epoch": 0.5189990732159406,
+ "grad_norm": 68.1931381225586,
+ "learning_rate": 1.5613382899628255e-05,
+ "loss": 10.0876,
+ "step": 140
+ },
+ {
+ "epoch": 0.5560704355885079,
+ "grad_norm": 49.315216064453125,
+ "learning_rate": 1.6728624535315986e-05,
+ "loss": 10.3721,
+ "step": 150
+ },
+ {
+ "epoch": 0.5560704355885079,
+ "eval_loss": 2.559400796890259,
+ "eval_runtime": 34.304,
+ "eval_samples_per_second": 47.166,
+ "eval_steps_per_second": 6.763,
+ "step": 150
+ },
+ {
+ "epoch": 0.593141797961075,
+ "grad_norm": 34.1899528503418,
+ "learning_rate": 1.7843866171003718e-05,
+ "loss": 10.4711,
+ "step": 160
+ },
+ {
+ "epoch": 0.6302131603336423,
+ "grad_norm": 37.074134826660156,
+ "learning_rate": 1.895910780669145e-05,
+ "loss": 9.9383,
+ "step": 170
+ },
+ {
+ "epoch": 0.6672845227062094,
+ "grad_norm": 49.323158264160156,
+ "learning_rate": 2.0074349442379185e-05,
+ "loss": 10.0448,
+ "step": 180
+ },
+ {
+ "epoch": 0.7043558850787767,
+ "grad_norm": 53.07639694213867,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 9.8467,
+ "step": 190
+ },
+ {
+ "epoch": 0.7414272474513438,
+ "grad_norm": 38.345916748046875,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 9.9434,
+ "step": 200
+ },
+ {
+ "epoch": 0.7784986098239111,
+ "grad_norm": 47.55274963378906,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 9.7912,
+ "step": 210
+ },
+ {
+ "epoch": 0.8155699721964782,
+ "grad_norm": 59.17176055908203,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 9.7905,
+ "step": 220
+ },
+ {
+ "epoch": 0.8526413345690455,
+ "grad_norm": 52.57457733154297,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 9.9412,
+ "step": 230
+ },
+ {
+ "epoch": 0.8897126969416126,
+ "grad_norm": 42.251304626464844,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 10.0084,
+ "step": 240
+ },
+ {
+ "epoch": 0.9267840593141798,
+ "grad_norm": 46.63966369628906,
+ "learning_rate": 2.788104089219331e-05,
+ "loss": 10.0139,
+ "step": 250
+ },
+ {
+ "epoch": 0.963855421686747,
+ "grad_norm": 27.772289276123047,
+ "learning_rate": 2.8996282527881043e-05,
+ "loss": 9.3794,
+ "step": 260
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 46.3179931640625,
+ "learning_rate": 2.9987608426270136e-05,
+ "loss": 9.5747,
+ "step": 270
+ },
+ {
+ "epoch": 1.0370713623725671,
+ "grad_norm": 45.746551513671875,
+ "learning_rate": 2.98636926889715e-05,
+ "loss": 9.6763,
+ "step": 280
+ },
+ {
+ "epoch": 1.0741427247451343,
+ "grad_norm": 49.50416564941406,
+ "learning_rate": 2.9739776951672864e-05,
+ "loss": 10.4789,
+ "step": 290
+ },
+ {
+ "epoch": 1.1112140871177016,
+ "grad_norm": 43.72733688354492,
+ "learning_rate": 2.9615861214374226e-05,
+ "loss": 9.7244,
+ "step": 300
+ },
+ {
+ "epoch": 1.1112140871177016,
+ "eval_loss": 2.6879734992980957,
+ "eval_runtime": 34.2522,
+ "eval_samples_per_second": 47.238,
+ "eval_steps_per_second": 6.773,
+ "step": 300
+ },
+ {
+ "epoch": 1.1482854494902688,
+ "grad_norm": 48.68376541137695,
+ "learning_rate": 2.9491945477075588e-05,
+ "loss": 9.9421,
+ "step": 310
+ },
+ {
+ "epoch": 1.185356811862836,
+ "grad_norm": 44.014747619628906,
+ "learning_rate": 2.9368029739776953e-05,
+ "loss": 9.5022,
+ "step": 320
+ },
+ {
+ "epoch": 1.222428174235403,
+ "grad_norm": 49.858642578125,
+ "learning_rate": 2.9244114002478315e-05,
+ "loss": 9.6801,
+ "step": 330
+ },
+ {
+ "epoch": 1.2594995366079704,
+ "grad_norm": 49.97459411621094,
+ "learning_rate": 2.9120198265179677e-05,
+ "loss": 9.25,
+ "step": 340
+ },
+ {
+ "epoch": 1.2965708989805376,
+ "grad_norm": 40.38698196411133,
+ "learning_rate": 2.8996282527881043e-05,
+ "loss": 10.2879,
+ "step": 350
+ },
+ {
+ "epoch": 1.3336422613531047,
+ "grad_norm": 29.797264099121094,
+ "learning_rate": 2.8872366790582405e-05,
+ "loss": 9.97,
+ "step": 360
+ },
+ {
+ "epoch": 1.3707136237256718,
+ "grad_norm": 43.83412551879883,
+ "learning_rate": 2.874845105328377e-05,
+ "loss": 9.3751,
+ "step": 370
+ },
+ {
+ "epoch": 1.407784986098239,
+ "grad_norm": 42.754154205322266,
+ "learning_rate": 2.8624535315985132e-05,
+ "loss": 9.8947,
+ "step": 380
+ },
+ {
+ "epoch": 1.4448563484708064,
+ "grad_norm": 50.08184051513672,
+ "learning_rate": 2.8500619578686494e-05,
+ "loss": 9.4987,
+ "step": 390
+ },
+ {
+ "epoch": 1.4819277108433735,
+ "grad_norm": 41.81029510498047,
+ "learning_rate": 2.8376703841387856e-05,
+ "loss": 10.1515,
+ "step": 400
+ },
+ {
+ "epoch": 1.5189990732159406,
+ "grad_norm": 44.03816223144531,
+ "learning_rate": 2.8252788104089218e-05,
+ "loss": 10.0098,
+ "step": 410
+ },
+ {
+ "epoch": 1.556070435588508,
+ "grad_norm": 42.871856689453125,
+ "learning_rate": 2.8128872366790583e-05,
+ "loss": 9.5566,
+ "step": 420
+ },
+ {
+ "epoch": 1.5931417979610751,
+ "grad_norm": 44.18041229248047,
+ "learning_rate": 2.8004956629491945e-05,
+ "loss": 10.1925,
+ "step": 430
+ },
+ {
+ "epoch": 1.6302131603336423,
+ "grad_norm": 30.012725830078125,
+ "learning_rate": 2.788104089219331e-05,
+ "loss": 9.7408,
+ "step": 440
+ },
+ {
+ "epoch": 1.6672845227062094,
+ "grad_norm": 38.104827880859375,
+ "learning_rate": 2.7757125154894673e-05,
+ "loss": 9.6873,
+ "step": 450
+ },
+ {
+ "epoch": 1.6672845227062094,
+ "eval_loss": 2.5061004161834717,
+ "eval_runtime": 34.2772,
+ "eval_samples_per_second": 47.203,
+ "eval_steps_per_second": 6.768,
+ "step": 450
+ },
+ {
+ "epoch": 1.7043558850787766,
+ "grad_norm": 42.93535232543945,
+ "learning_rate": 2.7633209417596035e-05,
+ "loss": 9.1932,
+ "step": 460
+ },
+ {
+ "epoch": 1.7414272474513437,
+ "grad_norm": 63.935279846191406,
+ "learning_rate": 2.75092936802974e-05,
+ "loss": 9.9204,
+ "step": 470
+ },
+ {
+ "epoch": 1.778498609823911,
+ "grad_norm": 30.906002044677734,
+ "learning_rate": 2.7385377942998762e-05,
+ "loss": 9.5151,
+ "step": 480
+ },
+ {
+ "epoch": 1.8155699721964782,
+ "grad_norm": 48.05679702758789,
+ "learning_rate": 2.7261462205700128e-05,
+ "loss": 9.8418,
+ "step": 490
+ },
+ {
+ "epoch": 1.8526413345690456,
+ "grad_norm": 56.36283874511719,
+ "learning_rate": 2.7137546468401486e-05,
+ "loss": 9.6943,
+ "step": 500
+ },
+ {
+ "epoch": 1.8897126969416127,
+ "grad_norm": 47.22933578491211,
+ "learning_rate": 2.7013630731102848e-05,
+ "loss": 9.6948,
+ "step": 510
+ },
+ {
+ "epoch": 1.9267840593141798,
+ "grad_norm": 36.110008239746094,
+ "learning_rate": 2.6889714993804214e-05,
+ "loss": 9.2274,
+ "step": 520
+ },
+ {
+ "epoch": 1.963855421686747,
+ "grad_norm": 37.13912582397461,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 10.138,
+ "step": 530
+ },
+ {
+ "epoch": 2.0,
+ "grad_norm": 51.37559509277344,
+ "learning_rate": 2.664188351920694e-05,
+ "loss": 8.7934,
+ "step": 540
+ },
+ {
+ "epoch": 2.037071362372567,
+ "grad_norm": 42.32170104980469,
+ "learning_rate": 2.6517967781908303e-05,
+ "loss": 8.8304,
+ "step": 550
+ },
+ {
+ "epoch": 2.0741427247451343,
+ "grad_norm": 42.406219482421875,
+ "learning_rate": 2.6394052044609665e-05,
+ "loss": 8.9593,
+ "step": 560
+ },
+ {
+ "epoch": 2.1112140871177014,
+ "grad_norm": 52.097023010253906,
+ "learning_rate": 2.627013630731103e-05,
+ "loss": 9.0051,
+ "step": 570
+ },
+ {
+ "epoch": 2.1482854494902686,
+ "grad_norm": 83.86861419677734,
+ "learning_rate": 2.6146220570012392e-05,
+ "loss": 9.2829,
+ "step": 580
+ },
+ {
+ "epoch": 2.185356811862836,
+ "grad_norm": 38.229949951171875,
+ "learning_rate": 2.6022304832713758e-05,
+ "loss": 9.785,
+ "step": 590
+ },
+ {
+ "epoch": 2.2224281742354033,
+ "grad_norm": 50.322513580322266,
+ "learning_rate": 2.5898389095415116e-05,
+ "loss": 9.6401,
+ "step": 600
+ },
+ {
+ "epoch": 2.2224281742354033,
+ "eval_loss": 2.448084592819214,
+ "eval_runtime": 34.2525,
+ "eval_samples_per_second": 47.237,
+ "eval_steps_per_second": 6.773,
+ "step": 600
+ },
+ {
+ "epoch": 2.266913809082484,
+ "grad_norm": 39.366676330566406,
+ "learning_rate": 2.5774473358116482e-05,
+ "loss": 8.7059,
+ "step": 610
+ },
+ {
+ "epoch": 2.303985171455051,
+ "grad_norm": 41.00213623046875,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 9.1285,
+ "step": 620
+ },
+ {
+ "epoch": 2.341056533827618,
+ "grad_norm": 63.442169189453125,
+ "learning_rate": 2.5526641883519206e-05,
+ "loss": 9.3277,
+ "step": 630
+ },
+ {
+ "epoch": 2.3781278962001853,
+ "grad_norm": 51.54194641113281,
+ "learning_rate": 2.540272614622057e-05,
+ "loss": 9.1577,
+ "step": 640
+ },
+ {
+ "epoch": 2.4151992585727524,
+ "grad_norm": 57.583885192871094,
+ "learning_rate": 2.5278810408921933e-05,
+ "loss": 9.6032,
+ "step": 650
+ },
+ {
+ "epoch": 2.4522706209453196,
+ "grad_norm": 60.13545227050781,
+ "learning_rate": 2.51548946716233e-05,
+ "loss": 8.5451,
+ "step": 660
+ },
+ {
+ "epoch": 2.489341983317887,
+ "grad_norm": 139.85794067382812,
+ "learning_rate": 2.503097893432466e-05,
+ "loss": 8.6601,
+ "step": 670
+ },
+ {
+ "epoch": 2.5264133456904543,
+ "grad_norm": 160.73171997070312,
+ "learning_rate": 2.4907063197026023e-05,
+ "loss": 9.522,
+ "step": 680
+ },
+ {
+ "epoch": 2.5634847080630214,
+ "grad_norm": 190.7194366455078,
+ "learning_rate": 2.4783147459727385e-05,
+ "loss": 10.0619,
+ "step": 690
+ },
+ {
+ "epoch": 2.6005560704355886,
+ "grad_norm": 172.5655517578125,
+ "learning_rate": 2.4659231722428747e-05,
+ "loss": 9.6444,
+ "step": 700
+ },
+ {
+ "epoch": 2.6376274328081557,
+ "grad_norm": 280.9183044433594,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 10.7593,
+ "step": 710
+ },
+ {
+ "epoch": 2.674698795180723,
+ "grad_norm": 118.03450775146484,
+ "learning_rate": 2.4411400247831474e-05,
+ "loss": 10.4432,
+ "step": 720
+ },
+ {
+ "epoch": 2.71177015755329,
+ "grad_norm": 332.662353515625,
+ "learning_rate": 2.428748451053284e-05,
+ "loss": 9.6033,
+ "step": 730
+ },
+ {
+ "epoch": 2.748841519925857,
+ "grad_norm": 129.18788146972656,
+ "learning_rate": 2.41635687732342e-05,
+ "loss": 9.108,
+ "step": 740
+ },
+ {
+ "epoch": 2.7859128822984243,
+ "grad_norm": 350.5856628417969,
+ "learning_rate": 2.4039653035935564e-05,
+ "loss": 9.884,
+ "step": 750
+ },
+ {
+ "epoch": 2.7859128822984243,
+ "eval_loss": 2.599517583847046,
+ "eval_runtime": 34.0592,
+ "eval_samples_per_second": 47.506,
+ "eval_steps_per_second": 6.812,
+ "step": 750
+ },
+ {
+ "epoch": 2.822984244670992,
+ "grad_norm": 285.14862060546875,
+ "learning_rate": 2.391573729863693e-05,
+ "loss": 11.3952,
+ "step": 760
+ },
+ {
+ "epoch": 2.860055607043559,
+ "grad_norm": 185.5862579345703,
+ "learning_rate": 2.379182156133829e-05,
+ "loss": 10.6798,
+ "step": 770
+ },
+ {
+ "epoch": 2.897126969416126,
+ "grad_norm": 77.20765686035156,
+ "learning_rate": 2.3667905824039656e-05,
+ "loss": 9.9795,
+ "step": 780
+ },
+ {
+ "epoch": 2.9341983317886933,
+ "grad_norm": 88.75109100341797,
+ "learning_rate": 2.3543990086741015e-05,
+ "loss": 9.3694,
+ "step": 790
+ },
+ {
+ "epoch": 2.9712696941612604,
+ "grad_norm": 53.416690826416016,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 8.4631,
+ "step": 800
+ },
+ {
+ "epoch": 3.0111214087117704,
+ "grad_norm": 72.12966918945312,
+ "learning_rate": 2.3296158612143742e-05,
+ "loss": 9.3903,
+ "step": 810
+ },
+ {
+ "epoch": 3.0481927710843375,
+ "grad_norm": 159.3909912109375,
+ "learning_rate": 2.3172242874845104e-05,
+ "loss": 8.9608,
+ "step": 820
+ },
+ {
+ "epoch": 3.0852641334569046,
+ "grad_norm": 161.03713989257812,
+ "learning_rate": 2.304832713754647e-05,
+ "loss": 9.623,
+ "step": 830
+ },
+ {
+ "epoch": 3.122335495829472,
+ "grad_norm": 136.50045776367188,
+ "learning_rate": 2.2924411400247832e-05,
+ "loss": 8.5958,
+ "step": 840
+ },
+ {
+ "epoch": 3.159406858202039,
+ "grad_norm": 60.785926818847656,
+ "learning_rate": 2.2800495662949197e-05,
+ "loss": 8.6278,
+ "step": 850
+ },
+ {
+ "epoch": 3.196478220574606,
+ "grad_norm": 115.2509536743164,
+ "learning_rate": 2.267657992565056e-05,
+ "loss": 8.661,
+ "step": 860
+ },
+ {
+ "epoch": 3.233549582947173,
+ "grad_norm": 144.38912963867188,
+ "learning_rate": 2.255266418835192e-05,
+ "loss": 9.5836,
+ "step": 870
+ },
+ {
+ "epoch": 3.2706209453197403,
+ "grad_norm": 239.74520874023438,
+ "learning_rate": 2.2428748451053287e-05,
+ "loss": 8.6221,
+ "step": 880
+ },
+ {
+ "epoch": 3.3076923076923075,
+ "grad_norm": 215.12789916992188,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 9.1995,
+ "step": 890
+ },
+ {
+ "epoch": 3.344763670064875,
+ "grad_norm": 113.77183532714844,
+ "learning_rate": 2.218091697645601e-05,
+ "loss": 8.6909,
+ "step": 900
+ },
+ {
+ "epoch": 3.344763670064875,
+ "eval_loss": 2.4246325492858887,
+ "eval_runtime": 34.1351,
+ "eval_samples_per_second": 47.4,
+ "eval_steps_per_second": 6.797,
+ "step": 900
+ },
+ {
+ "epoch": 3.381835032437442,
+ "grad_norm": 175.02731323242188,
+ "learning_rate": 2.2057001239157373e-05,
+ "loss": 8.7632,
+ "step": 910
+ },
+ {
+ "epoch": 3.4189063948100094,
+ "grad_norm": 142.76580810546875,
+ "learning_rate": 2.1933085501858735e-05,
+ "loss": 8.5604,
+ "step": 920
+ },
+ {
+ "epoch": 3.4559777571825765,
+ "grad_norm": 163.0391387939453,
+ "learning_rate": 2.18091697645601e-05,
+ "loss": 9.3871,
+ "step": 930
+ },
+ {
+ "epoch": 3.4930491195551436,
+ "grad_norm": 386.4294738769531,
+ "learning_rate": 2.1685254027261462e-05,
+ "loss": 9.6445,
+ "step": 940
+ },
+ {
+ "epoch": 3.5301204819277108,
+ "grad_norm": 86.92581939697266,
+ "learning_rate": 2.1561338289962827e-05,
+ "loss": 9.635,
+ "step": 950
+ },
+ {
+ "epoch": 3.567191844300278,
+ "grad_norm": 80.94845581054688,
+ "learning_rate": 2.143742255266419e-05,
+ "loss": 8.8689,
+ "step": 960
+ },
+ {
+ "epoch": 3.6042632066728455,
+ "grad_norm": 69.99708557128906,
+ "learning_rate": 2.1313506815365555e-05,
+ "loss": 9.1411,
+ "step": 970
+ },
+ {
+ "epoch": 3.641334569045412,
+ "grad_norm": 80.53229522705078,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 8.7552,
+ "step": 980
+ },
+ {
+ "epoch": 3.67840593141798,
+ "grad_norm": 53.3475341796875,
+ "learning_rate": 2.1065675340768275e-05,
+ "loss": 8.5697,
+ "step": 990
+ },
+ {
+ "epoch": 3.715477293790547,
+ "grad_norm": 54.19875717163086,
+ "learning_rate": 2.094175960346964e-05,
+ "loss": 8.9008,
+ "step": 1000
+ },
+ {
+ "epoch": 3.752548656163114,
+ "grad_norm": 52.96628952026367,
+ "learning_rate": 2.0817843866171003e-05,
+ "loss": 8.9415,
+ "step": 1010
+ },
+ {
+ "epoch": 3.789620018535681,
+ "grad_norm": 64.08312225341797,
+ "learning_rate": 2.0693928128872368e-05,
+ "loss": 8.1303,
+ "step": 1020
+ },
+ {
+ "epoch": 3.8266913809082483,
+ "grad_norm": 68.22163391113281,
+ "learning_rate": 2.057001239157373e-05,
+ "loss": 8.241,
+ "step": 1030
+ },
+ {
+ "epoch": 3.8637627432808155,
+ "grad_norm": 66.70902252197266,
+ "learning_rate": 2.0446096654275092e-05,
+ "loss": 8.6969,
+ "step": 1040
+ },
+ {
+ "epoch": 3.9008341056533826,
+ "grad_norm": 49.94658279418945,
+ "learning_rate": 2.0322180916976458e-05,
+ "loss": 9.8392,
+ "step": 1050
+ },
+ {
+ "epoch": 3.9008341056533826,
+ "eval_loss": 2.533297300338745,
+ "eval_runtime": 34.1106,
+ "eval_samples_per_second": 47.434,
+ "eval_steps_per_second": 6.801,
+ "step": 1050
+ },
+ {
+ "epoch": 3.93790546802595,
+ "grad_norm": 73.9527816772461,
+ "learning_rate": 2.6765799256505576e-05,
+ "loss": 8.5934,
+ "step": 1060
+ },
+ {
+ "epoch": 3.974976830398517,
+ "grad_norm": 76.90135955810547,
+ "learning_rate": 2.6703841387856257e-05,
+ "loss": 8.006,
+ "step": 1070
+ },
+ {
+ "epoch": 4.014828544949027,
+ "grad_norm": 85.86236572265625,
+ "learning_rate": 2.664188351920694e-05,
+ "loss": 9.0081,
+ "step": 1080
+ },
+ {
+ "epoch": 4.051899907321594,
+ "grad_norm": 106.57842254638672,
+ "learning_rate": 2.6579925650557622e-05,
+ "loss": 8.6706,
+ "step": 1090
+ },
+ {
+ "epoch": 4.088971269694161,
+ "grad_norm": 68.10344696044922,
+ "learning_rate": 2.6517967781908303e-05,
+ "loss": 9.6146,
+ "step": 1100
+ },
+ {
+ "epoch": 4.126042632066729,
+ "grad_norm": 67.02091217041016,
+ "learning_rate": 2.6456009913258984e-05,
+ "loss": 9.225,
+ "step": 1110
+ },
+ {
+ "epoch": 4.163113994439295,
+ "grad_norm": 67.66644287109375,
+ "learning_rate": 2.6394052044609665e-05,
+ "loss": 8.7522,
+ "step": 1120
+ },
+ {
+ "epoch": 4.200185356811863,
+ "grad_norm": 59.26814651489258,
+ "learning_rate": 2.633209417596035e-05,
+ "loss": 9.0221,
+ "step": 1130
+ },
+ {
+ "epoch": 4.23725671918443,
+ "grad_norm": 88.07310485839844,
+ "learning_rate": 2.627013630731103e-05,
+ "loss": 9.6458,
+ "step": 1140
+ },
+ {
+ "epoch": 4.274328081556997,
+ "grad_norm": 69.33895874023438,
+ "learning_rate": 2.620817843866171e-05,
+ "loss": 8.7692,
+ "step": 1150
+ },
+ {
+ "epoch": 4.311399443929565,
+ "grad_norm": 43.823951721191406,
+ "learning_rate": 2.6146220570012392e-05,
+ "loss": 9.2874,
+ "step": 1160
+ },
+ {
+ "epoch": 4.348470806302132,
+ "grad_norm": 63.590980529785156,
+ "learning_rate": 2.6084262701363077e-05,
+ "loss": 8.9276,
+ "step": 1170
+ },
+ {
+ "epoch": 4.385542168674699,
+ "grad_norm": 43.4519157409668,
+ "learning_rate": 2.6022304832713758e-05,
+ "loss": 8.7444,
+ "step": 1180
+ },
+ {
+ "epoch": 4.422613531047266,
+ "grad_norm": 72.1529312133789,
+ "learning_rate": 2.5960346964064435e-05,
+ "loss": 8.7265,
+ "step": 1190
+ },
+ {
+ "epoch": 4.459684893419833,
+ "grad_norm": 101.67008972167969,
+ "learning_rate": 2.5898389095415116e-05,
+ "loss": 8.7642,
+ "step": 1200
+ },
+ {
+ "epoch": 4.459684893419833,
+ "eval_loss": 2.6471314430236816,
+ "eval_runtime": 34.2723,
+ "eval_samples_per_second": 47.21,
+ "eval_steps_per_second": 6.769,
+ "step": 1200
+ },
+ {
+ "epoch": 4.4967562557924,
+ "grad_norm": 66.03606414794922,
+ "learning_rate": 2.5836431226765797e-05,
+ "loss": 8.8917,
+ "step": 1210
+ },
+ {
+ "epoch": 4.533827618164968,
+ "grad_norm": 60.01662826538086,
+ "learning_rate": 2.5774473358116482e-05,
+ "loss": 9.2155,
+ "step": 1220
+ },
+ {
+ "epoch": 4.570898980537534,
+ "grad_norm": 83.86510467529297,
+ "learning_rate": 2.5712515489467163e-05,
+ "loss": 8.6101,
+ "step": 1230
+ },
+ {
+ "epoch": 4.607970342910102,
+ "grad_norm": 64.35486602783203,
+ "learning_rate": 2.5650557620817844e-05,
+ "loss": 8.9904,
+ "step": 1240
+ },
+ {
+ "epoch": 4.645041705282669,
+ "grad_norm": 60.44017791748047,
+ "learning_rate": 2.5588599752168525e-05,
+ "loss": 9.3272,
+ "step": 1250
+ },
+ {
+ "epoch": 4.682113067655236,
+ "grad_norm": 56.501102447509766,
+ "learning_rate": 2.5526641883519206e-05,
+ "loss": 7.9367,
+ "step": 1260
+ },
+ {
+ "epoch": 4.719184430027804,
+ "grad_norm": 65.01017761230469,
+ "learning_rate": 2.546468401486989e-05,
+ "loss": 8.5891,
+ "step": 1270
+ },
+ {
+ "epoch": 4.7562557924003706,
+ "grad_norm": 63.42623519897461,
+ "learning_rate": 2.540272614622057e-05,
+ "loss": 8.6286,
+ "step": 1280
+ },
+ {
+ "epoch": 4.793327154772938,
+ "grad_norm": 58.96747589111328,
+ "learning_rate": 2.5340768277571252e-05,
+ "loss": 7.9982,
+ "step": 1290
+ },
+ {
+ "epoch": 4.830398517145505,
+ "grad_norm": 96.77368927001953,
+ "learning_rate": 2.5278810408921933e-05,
+ "loss": 7.5587,
+ "step": 1300
+ },
+ {
+ "epoch": 4.867469879518072,
+ "grad_norm": 85.92269897460938,
+ "learning_rate": 2.5216852540272614e-05,
+ "loss": 7.9405,
+ "step": 1310
+ },
+ {
+ "epoch": 4.904541241890639,
+ "grad_norm": 86.37085723876953,
+ "learning_rate": 2.51548946716233e-05,
+ "loss": 9.7092,
+ "step": 1320
+ },
+ {
+ "epoch": 4.941612604263207,
+ "grad_norm": 65.88274383544922,
+ "learning_rate": 2.509293680297398e-05,
+ "loss": 8.1475,
+ "step": 1330
+ },
+ {
+ "epoch": 4.978683966635774,
+ "grad_norm": 53.252742767333984,
+ "learning_rate": 2.503097893432466e-05,
+ "loss": 9.3603,
+ "step": 1340
+ },
+ {
+ "epoch": 5.014828544949027,
+ "grad_norm": 182.35781860351562,
+ "learning_rate": 2.4969021065675342e-05,
+ "loss": 7.6621,
+ "step": 1350
+ },
+ {
+ "epoch": 5.014828544949027,
+ "eval_loss": 2.8308892250061035,
+ "eval_runtime": 34.1063,
+ "eval_samples_per_second": 47.44,
+ "eval_steps_per_second": 6.802,
+ "step": 1350
+ },
+ {
+ "epoch": 5.051899907321594,
+ "grad_norm": 324.6659240722656,
+ "learning_rate": 2.4907063197026023e-05,
+ "loss": 9.2301,
+ "step": 1360
+ },
+ {
+ "epoch": 5.088971269694161,
+ "grad_norm": 4839.4453125,
+ "learning_rate": 2.4845105328376707e-05,
+ "loss": 9.7789,
+ "step": 1370
+ },
+ {
+ "epoch": 5.126042632066729,
+ "grad_norm": 408588.40625,
+ "learning_rate": 2.4783147459727385e-05,
+ "loss": 9.5359,
+ "step": 1380
+ },
+ {
+ "epoch": 5.163113994439295,
+ "grad_norm": 932.5144653320312,
+ "learning_rate": 2.4721189591078066e-05,
+ "loss": 10.8065,
+ "step": 1390
+ },
+ {
+ "epoch": 5.200185356811863,
+ "grad_norm": 148.0634765625,
+ "learning_rate": 2.4659231722428747e-05,
+ "loss": 10.0149,
+ "step": 1400
+ },
+ {
+ "epoch": 5.23725671918443,
+ "grad_norm": 1214.7279052734375,
+ "learning_rate": 2.459727385377943e-05,
+ "loss": 10.2582,
+ "step": 1410
+ },
+ {
+ "epoch": 5.274328081556997,
+ "grad_norm": 188.14292907714844,
+ "learning_rate": 2.4535315985130112e-05,
+ "loss": 10.16,
+ "step": 1420
+ },
+ {
+ "epoch": 5.311399443929565,
+ "grad_norm": 388.7134094238281,
+ "learning_rate": 2.4473358116480793e-05,
+ "loss": 10.0763,
+ "step": 1430
+ },
+ {
+ "epoch": 5.348470806302132,
+ "grad_norm": 270.61279296875,
+ "learning_rate": 2.4411400247831474e-05,
+ "loss": 9.5737,
+ "step": 1440
+ },
+ {
+ "epoch": 5.385542168674699,
+ "grad_norm": 1637.854736328125,
+ "learning_rate": 2.4349442379182155e-05,
+ "loss": 10.4816,
+ "step": 1450
+ },
+ {
+ "epoch": 5.422613531047266,
+ "grad_norm": 1118.7012939453125,
+ "learning_rate": 2.428748451053284e-05,
+ "loss": 8.6687,
+ "step": 1460
+ },
+ {
+ "epoch": 5.459684893419833,
+ "grad_norm": 788.9707641601562,
+ "learning_rate": 2.422552664188352e-05,
+ "loss": 8.4066,
+ "step": 1470
+ },
+ {
+ "epoch": 5.4967562557924,
+ "grad_norm": 185.22950744628906,
+ "learning_rate": 2.41635687732342e-05,
+ "loss": 9.386,
+ "step": 1480
+ },
+ {
+ "epoch": 5.533827618164968,
+ "grad_norm": 257.726806640625,
+ "learning_rate": 2.4101610904584883e-05,
+ "loss": 8.3911,
+ "step": 1490
+ },
+ {
+ "epoch": 5.570898980537534,
+ "grad_norm": 867.30908203125,
+ "learning_rate": 2.4039653035935564e-05,
+ "loss": 8.8025,
+ "step": 1500
+ },
+ {
+ "epoch": 5.570898980537534,
+ "eval_loss": 2.5407650470733643,
+ "eval_runtime": 34.0934,
+ "eval_samples_per_second": 47.458,
+ "eval_steps_per_second": 6.805,
+ "step": 1500
+ },
+ {
+ "epoch": 5.607970342910102,
+ "grad_norm": 68.35711669921875,
+ "learning_rate": 2.3977695167286248e-05,
+ "loss": 8.7939,
+ "step": 1510
+ },
+ {
+ "epoch": 5.645041705282669,
+ "grad_norm": 1419.1650390625,
+ "learning_rate": 2.391573729863693e-05,
+ "loss": 9.0903,
+ "step": 1520
+ },
+ {
+ "epoch": 5.682113067655236,
+ "grad_norm": 60.88732147216797,
+ "learning_rate": 2.385377942998761e-05,
+ "loss": 8.9878,
+ "step": 1530
+ },
+ {
+ "epoch": 5.719184430027804,
+ "grad_norm": 61.50148010253906,
+ "learning_rate": 2.379182156133829e-05,
+ "loss": 8.8642,
+ "step": 1540
+ },
+ {
+ "epoch": 5.7562557924003706,
+ "grad_norm": 73.9307861328125,
+ "learning_rate": 2.3729863692688972e-05,
+ "loss": 8.8625,
+ "step": 1550
+ },
+ {
+ "epoch": 5.793327154772938,
+ "grad_norm": 82.28329467773438,
+ "learning_rate": 2.3667905824039656e-05,
+ "loss": 8.4105,
+ "step": 1560
+ },
+ {
+ "epoch": 5.830398517145505,
+ "grad_norm": 83.54141235351562,
+ "learning_rate": 2.3605947955390337e-05,
+ "loss": 9.0163,
+ "step": 1570
+ },
+ {
+ "epoch": 5.867469879518072,
+ "grad_norm": 66.59397888183594,
+ "learning_rate": 2.3543990086741015e-05,
+ "loss": 8.8947,
+ "step": 1580
+ },
+ {
+ "epoch": 5.904541241890639,
+ "grad_norm": 55.8841667175293,
+ "learning_rate": 2.3482032218091696e-05,
+ "loss": 8.5647,
+ "step": 1590
+ },
+ {
+ "epoch": 5.941612604263207,
+ "grad_norm": 74.56892395019531,
+ "learning_rate": 2.3420074349442377e-05,
+ "loss": 7.7047,
+ "step": 1600
+ },
+ {
+ "epoch": 5.978683966635774,
+ "grad_norm": 106.54405212402344,
+ "learning_rate": 2.335811648079306e-05,
+ "loss": 8.1484,
+ "step": 1610
+ },
+ {
+ "epoch": 6.014828544949027,
+ "grad_norm": 107.8216781616211,
+ "learning_rate": 2.3296158612143742e-05,
+ "loss": 8.4079,
+ "step": 1620
+ },
+ {
+ "epoch": 6.051899907321594,
+ "grad_norm": 100.98133087158203,
+ "learning_rate": 2.3234200743494423e-05,
+ "loss": 8.5027,
+ "step": 1630
+ },
+ {
+ "epoch": 6.088971269694161,
+ "grad_norm": 67.31327819824219,
+ "learning_rate": 2.3172242874845104e-05,
+ "loss": 8.1805,
+ "step": 1640
+ },
+ {
+ "epoch": 6.126042632066729,
+ "grad_norm": 61.10900115966797,
+ "learning_rate": 2.311028500619579e-05,
+ "loss": 8.4519,
+ "step": 1650
+ },
+ {
+ "epoch": 6.126042632066729,
+ "eval_loss": 2.590090036392212,
+ "eval_runtime": 34.1392,
+ "eval_samples_per_second": 47.394,
+ "eval_steps_per_second": 6.796,
+ "step": 1650
+ },
+ {
+ "epoch": 6.163113994439295,
+ "grad_norm": 60.91183090209961,
+ "learning_rate": 2.304832713754647e-05,
+ "loss": 9.062,
+ "step": 1660
+ },
+ {
+ "epoch": 6.200185356811863,
+ "grad_norm": 76.2873306274414,
+ "learning_rate": 2.298636926889715e-05,
+ "loss": 8.8499,
+ "step": 1670
+ },
+ {
+ "epoch": 6.23725671918443,
+ "grad_norm": 74.2361068725586,
+ "learning_rate": 2.2924411400247832e-05,
+ "loss": 8.6576,
+ "step": 1680
+ },
+ {
+ "epoch": 6.274328081556997,
+ "grad_norm": 84.99951934814453,
+ "learning_rate": 2.2862453531598513e-05,
+ "loss": 8.4652,
+ "step": 1690
+ },
+ {
+ "epoch": 6.311399443929565,
+ "grad_norm": 86.40947723388672,
+ "learning_rate": 2.2800495662949197e-05,
+ "loss": 9.0782,
+ "step": 1700
+ },
+ {
+ "epoch": 6.348470806302132,
+ "grad_norm": 75.45568084716797,
+ "learning_rate": 2.2738537794299878e-05,
+ "loss": 8.1532,
+ "step": 1710
+ },
+ {
+ "epoch": 6.385542168674699,
+ "grad_norm": 86.65908813476562,
+ "learning_rate": 2.267657992565056e-05,
+ "loss": 8.5185,
+ "step": 1720
+ },
+ {
+ "epoch": 6.422613531047266,
+ "grad_norm": 62.620758056640625,
+ "learning_rate": 2.261462205700124e-05,
+ "loss": 9.5908,
+ "step": 1730
+ },
+ {
+ "epoch": 6.459684893419833,
+ "grad_norm": 73.63095092773438,
+ "learning_rate": 2.255266418835192e-05,
+ "loss": 8.4188,
+ "step": 1740
+ },
+ {
+ "epoch": 6.4967562557924,
+ "grad_norm": 55.175045013427734,
+ "learning_rate": 2.2490706319702606e-05,
+ "loss": 8.1885,
+ "step": 1750
+ },
+ {
+ "epoch": 6.533827618164968,
+ "grad_norm": 47.647621154785156,
+ "learning_rate": 2.2428748451053287e-05,
+ "loss": 8.7666,
+ "step": 1760
+ },
+ {
+ "epoch": 6.570898980537534,
+ "grad_norm": 191.0419158935547,
+ "learning_rate": 2.2366790582403964e-05,
+ "loss": 8.6105,
+ "step": 1770
+ },
+ {
+ "epoch": 6.607970342910102,
+ "grad_norm": 57.81939697265625,
+ "learning_rate": 2.2304832713754645e-05,
+ "loss": 8.664,
+ "step": 1780
+ },
+ {
+ "epoch": 6.645041705282669,
+ "grad_norm": 85.04927825927734,
+ "learning_rate": 2.2242874845105326e-05,
+ "loss": 8.5294,
+ "step": 1790
+ },
+ {
+ "epoch": 6.682113067655236,
+ "grad_norm": 55.604122161865234,
+ "learning_rate": 2.218091697645601e-05,
+ "loss": 9.1857,
+ "step": 1800
+ },
+ {
+ "epoch": 6.682113067655236,
+ "eval_loss": 2.497358560562134,
+ "eval_runtime": 34.0441,
+ "eval_samples_per_second": 47.527,
+ "eval_steps_per_second": 6.815,
+ "step": 1800
+ },
+ {
+ "epoch": 6.719184430027804,
+ "grad_norm": 75.21812438964844,
+ "learning_rate": 2.211895910780669e-05,
+ "loss": 8.7053,
+ "step": 1810
+ },
+ {
+ "epoch": 6.7562557924003706,
+ "grad_norm": 85.76676940917969,
+ "learning_rate": 2.2057001239157373e-05,
+ "loss": 8.1428,
+ "step": 1820
+ },
+ {
+ "epoch": 6.793327154772938,
+ "grad_norm": 69.85575866699219,
+ "learning_rate": 2.1995043370508054e-05,
+ "loss": 8.4988,
+ "step": 1830
+ },
+ {
+ "epoch": 6.830398517145505,
+ "grad_norm": 66.85325622558594,
+ "learning_rate": 2.1933085501858735e-05,
+ "loss": 8.4147,
+ "step": 1840
+ },
+ {
+ "epoch": 6.867469879518072,
+ "grad_norm": 60.645294189453125,
+ "learning_rate": 2.187112763320942e-05,
+ "loss": 9.069,
+ "step": 1850
+ },
+ {
+ "epoch": 6.904541241890639,
+ "grad_norm": 77.59921264648438,
+ "learning_rate": 2.18091697645601e-05,
+ "loss": 8.4405,
+ "step": 1860
+ },
+ {
+ "epoch": 6.941612604263207,
+ "grad_norm": 56.27996826171875,
+ "learning_rate": 2.174721189591078e-05,
+ "loss": 9.2157,
+ "step": 1870
+ },
+ {
+ "epoch": 6.978683966635774,
+ "grad_norm": 64.18400573730469,
+ "learning_rate": 2.1685254027261462e-05,
+ "loss": 9.5492,
+ "step": 1880
+ },
+ {
+ "epoch": 7.014828544949027,
+ "grad_norm": 76.78484344482422,
+ "learning_rate": 2.1623296158612143e-05,
+ "loss": 8.1325,
+ "step": 1890
+ },
+ {
+ "epoch": 7.051899907321594,
+ "grad_norm": 77.69871520996094,
+ "learning_rate": 2.1561338289962827e-05,
+ "loss": 8.324,
+ "step": 1900
+ },
+ {
+ "epoch": 7.088971269694161,
+ "grad_norm": 102.43340301513672,
+ "learning_rate": 2.149938042131351e-05,
+ "loss": 7.7097,
+ "step": 1910
+ },
+ {
+ "epoch": 7.126042632066729,
+ "grad_norm": 114.0189208984375,
+ "learning_rate": 2.143742255266419e-05,
+ "loss": 8.0982,
+ "step": 1920
+ },
+ {
+ "epoch": 7.163113994439295,
+ "grad_norm": 66.60838317871094,
+ "learning_rate": 2.137546468401487e-05,
+ "loss": 7.7669,
+ "step": 1930
+ },
+ {
+ "epoch": 7.200185356811863,
+ "grad_norm": 252.72503662109375,
+ "learning_rate": 2.1313506815365555e-05,
+ "loss": 7.809,
+ "step": 1940
+ },
+ {
+ "epoch": 7.23725671918443,
+ "grad_norm": 91.0379409790039,
+ "learning_rate": 2.1251548946716236e-05,
+ "loss": 7.9729,
+ "step": 1950
+ },
+ {
+ "epoch": 7.23725671918443,
+ "eval_loss": 2.610764741897583,
+ "eval_runtime": 34.0774,
+ "eval_samples_per_second": 47.48,
+ "eval_steps_per_second": 6.808,
+ "step": 1950
+ },
+ {
+ "epoch": 7.274328081556997,
+ "grad_norm": 96.09555053710938,
+ "learning_rate": 2.1189591078066917e-05,
+ "loss": 8.2125,
+ "step": 1960
+ },
+ {
+ "epoch": 7.311399443929565,
+ "grad_norm": 68.51132202148438,
+ "learning_rate": 2.1127633209417594e-05,
+ "loss": 7.7403,
+ "step": 1970
+ },
+ {
+ "epoch": 7.348470806302132,
+ "grad_norm": 94.45308685302734,
+ "learning_rate": 2.1065675340768275e-05,
+ "loss": 7.5494,
+ "step": 1980
+ },
+ {
+ "epoch": 7.385542168674699,
+ "grad_norm": 56.785587310791016,
+ "learning_rate": 2.100371747211896e-05,
+ "loss": 8.2821,
+ "step": 1990
+ },
+ {
+ "epoch": 7.422613531047266,
+ "grad_norm": 91.59173583984375,
+ "learning_rate": 2.094175960346964e-05,
+ "loss": 8.1644,
+ "step": 2000
+ },
+ {
+ "epoch": 7.459684893419833,
+ "grad_norm": 78.57572174072266,
+ "learning_rate": 2.0879801734820322e-05,
+ "loss": 8.1664,
+ "step": 2010
+ },
+ {
+ "epoch": 7.4967562557924,
+ "grad_norm": 71.83062744140625,
+ "learning_rate": 2.0817843866171003e-05,
+ "loss": 8.5876,
+ "step": 2020
+ },
+ {
+ "epoch": 7.533827618164968,
+ "grad_norm": 74.75129699707031,
+ "learning_rate": 2.0755885997521684e-05,
+ "loss": 8.2753,
+ "step": 2030
+ },
+ {
+ "epoch": 7.570898980537534,
+ "grad_norm": 96.26950073242188,
+ "learning_rate": 2.0693928128872368e-05,
+ "loss": 9.2057,
+ "step": 2040
+ },
+ {
+ "epoch": 7.607970342910102,
+ "grad_norm": 96.87144470214844,
+ "learning_rate": 2.063197026022305e-05,
+ "loss": 8.0052,
+ "step": 2050
+ },
+ {
+ "epoch": 7.645041705282669,
+ "grad_norm": 83.44422912597656,
+ "learning_rate": 2.057001239157373e-05,
+ "loss": 8.4954,
+ "step": 2060
+ },
+ {
+ "epoch": 7.682113067655236,
+ "grad_norm": 86.1534194946289,
+ "learning_rate": 2.050805452292441e-05,
+ "loss": 8.0325,
+ "step": 2070
+ },
+ {
+ "epoch": 7.719184430027804,
+ "grad_norm": 89.11744689941406,
+ "learning_rate": 2.0446096654275092e-05,
+ "loss": 8.2934,
+ "step": 2080
+ },
+ {
+ "epoch": 7.7562557924003706,
+ "grad_norm": 60.81666946411133,
+ "learning_rate": 2.0384138785625777e-05,
+ "loss": 9.4019,
+ "step": 2090
+ },
+ {
+ "epoch": 7.793327154772938,
+ "grad_norm": 82.93228912353516,
+ "learning_rate": 2.0322180916976458e-05,
+ "loss": 8.874,
+ "step": 2100
+ },
+ {
+ "epoch": 7.793327154772938,
+ "eval_loss": 2.452894926071167,
+ "eval_runtime": 34.0684,
+ "eval_samples_per_second": 47.493,
+ "eval_steps_per_second": 6.81,
+ "step": 2100
+ },
+ {
+ "epoch": 7.830398517145505,
+ "grad_norm": 62.004817962646484,
+ "learning_rate": 2.026022304832714e-05,
+ "loss": 8.0985,
+ "step": 2110
+ },
+ {
+ "epoch": 7.867469879518072,
+ "grad_norm": 72.79752349853516,
+ "learning_rate": 2.019826517967782e-05,
+ "loss": 8.9063,
+ "step": 2120
+ },
+ {
+ "epoch": 7.904541241890639,
+ "grad_norm": 1507.2855224609375,
+ "learning_rate": 2.01363073110285e-05,
+ "loss": 8.6332,
+ "step": 2130
+ },
+ {
+ "epoch": 7.941612604263207,
+ "grad_norm": 134.11378479003906,
+ "learning_rate": 2.0074349442379185e-05,
+ "loss": 9.4106,
+ "step": 2140
+ },
+ {
+ "epoch": 7.978683966635774,
+ "grad_norm": 72.19046783447266,
+ "learning_rate": 2.0012391573729866e-05,
+ "loss": 8.2304,
+ "step": 2150
+ },
+ {
+ "epoch": 8.014828544949028,
+ "grad_norm": 57.130165100097656,
+ "learning_rate": 1.9950433705080544e-05,
+ "loss": 7.9675,
+ "step": 2160
+ },
+ {
+ "epoch": 8.051899907321594,
+ "grad_norm": 72.58036041259766,
+ "learning_rate": 1.9888475836431225e-05,
+ "loss": 7.602,
+ "step": 2170
+ },
+ {
+ "epoch": 8.088971269694161,
+ "grad_norm": 91.77513885498047,
+ "learning_rate": 1.982651796778191e-05,
+ "loss": 7.9151,
+ "step": 2180
+ },
+ {
+ "epoch": 8.126042632066728,
+ "grad_norm": 111.52696228027344,
+ "learning_rate": 1.976456009913259e-05,
+ "loss": 7.5653,
+ "step": 2190
+ },
+ {
+ "epoch": 8.163113994439296,
+ "grad_norm": 74.9671401977539,
+ "learning_rate": 1.970260223048327e-05,
+ "loss": 7.1567,
+ "step": 2200
+ },
+ {
+ "epoch": 8.200185356811863,
+ "grad_norm": 124.09498596191406,
+ "learning_rate": 1.9640644361833952e-05,
+ "loss": 8.1708,
+ "step": 2210
+ },
+ {
+ "epoch": 8.23725671918443,
+ "grad_norm": 95.21754455566406,
+ "learning_rate": 1.9578686493184633e-05,
+ "loss": 7.0013,
+ "step": 2220
+ },
+ {
+ "epoch": 8.274328081556996,
+ "grad_norm": 135.3839569091797,
+ "learning_rate": 1.9516728624535318e-05,
+ "loss": 7.5999,
+ "step": 2230
+ },
+ {
+ "epoch": 8.311399443929565,
+ "grad_norm": 89.70751190185547,
+ "learning_rate": 1.9454770755886e-05,
+ "loss": 7.2028,
+ "step": 2240
+ },
+ {
+ "epoch": 8.348470806302132,
+ "grad_norm": 70.08137512207031,
+ "learning_rate": 1.939281288723668e-05,
+ "loss": 8.3759,
+ "step": 2250
+ },
+ {
+ "epoch": 8.348470806302132,
+ "eval_loss": 2.746619939804077,
+ "eval_runtime": 34.0831,
+ "eval_samples_per_second": 47.472,
+ "eval_steps_per_second": 6.807,
+ "step": 2250
+ },
+ {
+ "epoch": 8.400370713623726,
+ "grad_norm": 69.07305145263672,
+ "learning_rate": 1.933085501858736e-05,
+ "loss": 8.9083,
+ "step": 2260
+ },
+ {
+ "epoch": 8.437442075996293,
+ "grad_norm": 66.43407440185547,
+ "learning_rate": 1.926889714993804e-05,
+ "loss": 6.9349,
+ "step": 2270
+ },
+ {
+ "epoch": 8.47451343836886,
+ "grad_norm": 167.85784912109375,
+ "learning_rate": 1.9206939281288726e-05,
+ "loss": 7.5041,
+ "step": 2280
+ },
+ {
+ "epoch": 8.511584800741428,
+ "grad_norm": 173.38304138183594,
+ "learning_rate": 1.9144981412639407e-05,
+ "loss": 7.3744,
+ "step": 2290
+ },
+ {
+ "epoch": 8.548656163113995,
+ "grad_norm": 80.71511840820312,
+ "learning_rate": 1.9083023543990088e-05,
+ "loss": 8.6541,
+ "step": 2300
+ },
+ {
+ "epoch": 8.585727525486561,
+ "grad_norm": 82.77088928222656,
+ "learning_rate": 1.902106567534077e-05,
+ "loss": 8.6305,
+ "step": 2310
+ },
+ {
+ "epoch": 8.62279888785913,
+ "grad_norm": 55.3135871887207,
+ "learning_rate": 1.895910780669145e-05,
+ "loss": 8.2577,
+ "step": 2320
+ },
+ {
+ "epoch": 8.659870250231696,
+ "grad_norm": 64.61204528808594,
+ "learning_rate": 1.8897149938042134e-05,
+ "loss": 7.6382,
+ "step": 2330
+ },
+ {
+ "epoch": 8.696941612604263,
+ "grad_norm": 77.9918212890625,
+ "learning_rate": 1.8835192069392815e-05,
+ "loss": 8.114,
+ "step": 2340
+ },
+ {
+ "epoch": 8.73401297497683,
+ "grad_norm": 50.139678955078125,
+ "learning_rate": 1.8773234200743496e-05,
+ "loss": 7.8875,
+ "step": 2350
+ },
+ {
+ "epoch": 8.771084337349398,
+ "grad_norm": 241.43588256835938,
+ "learning_rate": 1.8711276332094174e-05,
+ "loss": 7.0444,
+ "step": 2360
+ },
+ {
+ "epoch": 8.808155699721965,
+ "grad_norm": 133.25645446777344,
+ "learning_rate": 1.8649318463444855e-05,
+ "loss": 7.7393,
+ "step": 2370
+ },
+ {
+ "epoch": 8.845227062094532,
+ "grad_norm": 94.1192855834961,
+ "learning_rate": 1.858736059479554e-05,
+ "loss": 8.8284,
+ "step": 2380
+ },
+ {
+ "epoch": 8.882298424467098,
+ "grad_norm": 83.68805694580078,
+ "learning_rate": 1.852540272614622e-05,
+ "loss": 7.997,
+ "step": 2390
+ },
+ {
+ "epoch": 8.919369786839667,
+ "grad_norm": 67.01426696777344,
+ "learning_rate": 1.84634448574969e-05,
+ "loss": 7.786,
+ "step": 2400
+ },
+ {
+ "epoch": 8.919369786839667,
+ "eval_loss": 2.6791422367095947,
+ "eval_runtime": 34.2417,
+ "eval_samples_per_second": 47.252,
+ "eval_steps_per_second": 6.775,
+ "step": 2400
+ },
+ {
+ "epoch": 8.956441149212234,
+ "grad_norm": 72.36073303222656,
+ "learning_rate": 1.8401486988847582e-05,
+ "loss": 7.6257,
+ "step": 2410
+ },
+ {
+ "epoch": 8.9935125115848,
+ "grad_norm": 70.3509521484375,
+ "learning_rate": 1.8339529120198267e-05,
+ "loss": 7.099,
+ "step": 2420
+ },
+ {
+ "epoch": 9.03336422613531,
+ "grad_norm": 136.34727478027344,
+ "learning_rate": 1.8277571251548948e-05,
+ "loss": 8.041,
+ "step": 2430
+ },
+ {
+ "epoch": 9.070435588507877,
+ "grad_norm": 81.48332214355469,
+ "learning_rate": 1.821561338289963e-05,
+ "loss": 7.8606,
+ "step": 2440
+ },
+ {
+ "epoch": 9.107506950880445,
+ "grad_norm": 93.1724853515625,
+ "learning_rate": 1.815365551425031e-05,
+ "loss": 7.8551,
+ "step": 2450
+ },
+ {
+ "epoch": 9.144578313253012,
+ "grad_norm": 132.82778930664062,
+ "learning_rate": 1.809169764560099e-05,
+ "loss": 7.3977,
+ "step": 2460
+ },
+ {
+ "epoch": 9.181649675625579,
+ "grad_norm": 102.73971557617188,
+ "learning_rate": 1.8029739776951675e-05,
+ "loss": 7.8721,
+ "step": 2470
+ },
+ {
+ "epoch": 9.218721037998147,
+ "grad_norm": 106.7470932006836,
+ "learning_rate": 1.7967781908302356e-05,
+ "loss": 7.5839,
+ "step": 2480
+ },
+ {
+ "epoch": 9.255792400370714,
+ "grad_norm": 84.64981842041016,
+ "learning_rate": 1.7905824039653037e-05,
+ "loss": 7.1823,
+ "step": 2490
+ },
+ {
+ "epoch": 9.29286376274328,
+ "grad_norm": 85.59486389160156,
+ "learning_rate": 1.7843866171003718e-05,
+ "loss": 7.5513,
+ "step": 2500
+ },
+ {
+ "epoch": 9.329935125115847,
+ "grad_norm": 89.53803253173828,
+ "learning_rate": 1.77819083023544e-05,
+ "loss": 8.0879,
+ "step": 2510
+ },
+ {
+ "epoch": 9.367006487488416,
+ "grad_norm": 43.886898040771484,
+ "learning_rate": 1.7719950433705084e-05,
+ "loss": 7.5694,
+ "step": 2520
+ },
+ {
+ "epoch": 9.404077849860982,
+ "grad_norm": 73.4898910522461,
+ "learning_rate": 1.7657992565055765e-05,
+ "loss": 7.3436,
+ "step": 2530
+ },
+ {
+ "epoch": 9.44114921223355,
+ "grad_norm": 100.42314147949219,
+ "learning_rate": 1.7596034696406446e-05,
+ "loss": 6.9425,
+ "step": 2540
+ },
+ {
+ "epoch": 9.478220574606116,
+ "grad_norm": 121.53675079345703,
+ "learning_rate": 1.7534076827757123e-05,
+ "loss": 7.9461,
+ "step": 2550
+ },
+ {
+ "epoch": 9.478220574606116,
+ "eval_loss": 2.6608545780181885,
+ "eval_runtime": 34.2437,
+ "eval_samples_per_second": 47.25,
+ "eval_steps_per_second": 6.775,
+ "step": 2550
+ },
+ {
+ "epoch": 9.515291936978684,
+ "grad_norm": 83.28184509277344,
+ "learning_rate": 1.7472118959107804e-05,
+ "loss": 6.782,
+ "step": 2560
+ },
+ {
+ "epoch": 9.552363299351251,
+ "grad_norm": 81.46753692626953,
+ "learning_rate": 1.741016109045849e-05,
+ "loss": 7.3027,
+ "step": 2570
+ },
+ {
+ "epoch": 9.589434661723818,
+ "grad_norm": 77.35752868652344,
+ "learning_rate": 1.734820322180917e-05,
+ "loss": 7.3348,
+ "step": 2580
+ },
+ {
+ "epoch": 9.626506024096386,
+ "grad_norm": 77.17479705810547,
+ "learning_rate": 1.728624535315985e-05,
+ "loss": 7.7864,
+ "step": 2590
+ },
+ {
+ "epoch": 9.663577386468953,
+ "grad_norm": 103.42965698242188,
+ "learning_rate": 1.722428748451053e-05,
+ "loss": 6.3552,
+ "step": 2600
+ },
+ {
+ "epoch": 9.70064874884152,
+ "grad_norm": 103.3230972290039,
+ "learning_rate": 1.7162329615861213e-05,
+ "loss": 7.151,
+ "step": 2610
+ },
+ {
+ "epoch": 9.737720111214086,
+ "grad_norm": 91.23753356933594,
+ "learning_rate": 1.7100371747211897e-05,
+ "loss": 6.1664,
+ "step": 2620
+ },
+ {
+ "epoch": 9.774791473586655,
+ "grad_norm": 68.76666259765625,
+ "learning_rate": 1.7038413878562578e-05,
+ "loss": 6.0398,
+ "step": 2630
+ },
+ {
+ "epoch": 9.811862835959221,
+ "grad_norm": 112.56900787353516,
+ "learning_rate": 1.697645600991326e-05,
+ "loss": 7.0452,
+ "step": 2640
+ },
+ {
+ "epoch": 9.848934198331788,
+ "grad_norm": 142.0828399658203,
+ "learning_rate": 1.691449814126394e-05,
+ "loss": 7.2457,
+ "step": 2650
+ },
+ {
+ "epoch": 9.886005560704357,
+ "grad_norm": 149.853515625,
+ "learning_rate": 1.685254027261462e-05,
+ "loss": 6.7531,
+ "step": 2660
+ },
+ {
+ "epoch": 9.923076923076923,
+ "grad_norm": 107.76028442382812,
+ "learning_rate": 1.6790582403965305e-05,
+ "loss": 6.7149,
+ "step": 2670
+ },
+ {
+ "epoch": 9.96014828544949,
+ "grad_norm": 83.82595825195312,
+ "learning_rate": 1.6728624535315986e-05,
+ "loss": 6.4635,
+ "step": 2680
+ },
+ {
+ "epoch": 9.997219647822057,
+ "grad_norm": 106.68073272705078,
+ "learning_rate": 1.6666666666666667e-05,
+ "loss": 6.2237,
+ "step": 2690
+ },
+ {
+ "epoch": 10.037071362372567,
+ "grad_norm": 78.0780029296875,
+ "learning_rate": 1.660470879801735e-05,
+ "loss": 6.1798,
+ "step": 2700
+ },
+ {
+ "epoch": 10.037071362372567,
+ "eval_loss": 2.993870735168457,
+ "eval_runtime": 34.3527,
+ "eval_samples_per_second": 47.1,
+ "eval_steps_per_second": 6.753,
+ "step": 2700
+ },
+ {
+ "epoch": 10.074142724745135,
+ "grad_norm": 177.5696563720703,
+ "learning_rate": 1.6542750929368033e-05,
+ "loss": 7.2224,
+ "step": 2710
+ },
+ {
+ "epoch": 10.111214087117702,
+ "grad_norm": 134.7563018798828,
+ "learning_rate": 1.6480793060718714e-05,
+ "loss": 6.5327,
+ "step": 2720
+ },
+ {
+ "epoch": 10.148285449490269,
+ "grad_norm": 122.22810363769531,
+ "learning_rate": 1.6418835192069395e-05,
+ "loss": 7.4686,
+ "step": 2730
+ },
+ {
+ "epoch": 10.185356811862835,
+ "grad_norm": 137.185302734375,
+ "learning_rate": 1.6356877323420076e-05,
+ "loss": 6.1404,
+ "step": 2740
+ },
+ {
+ "epoch": 10.222428174235404,
+ "grad_norm": 100.84989166259766,
+ "learning_rate": 1.6294919454770753e-05,
+ "loss": 7.0005,
+ "step": 2750
+ },
+ {
+ "epoch": 10.25949953660797,
+ "grad_norm": 122.00653839111328,
+ "learning_rate": 1.6232961586121438e-05,
+ "loss": 5.7726,
+ "step": 2760
+ },
+ {
+ "epoch": 10.296570898980537,
+ "grad_norm": 172.2955780029297,
+ "learning_rate": 1.617100371747212e-05,
+ "loss": 6.5327,
+ "step": 2770
+ },
+ {
+ "epoch": 10.333642261353106,
+ "grad_norm": 195.36851501464844,
+ "learning_rate": 1.61090458488228e-05,
+ "loss": 7.5015,
+ "step": 2780
+ },
+ {
+ "epoch": 10.370713623725672,
+ "grad_norm": 135.3350372314453,
+ "learning_rate": 1.604708798017348e-05,
+ "loss": 6.5526,
+ "step": 2790
+ },
+ {
+ "epoch": 10.407784986098239,
+ "grad_norm": 106.01554870605469,
+ "learning_rate": 1.5985130111524162e-05,
+ "loss": 6.2078,
+ "step": 2800
+ },
+ {
+ "epoch": 10.444856348470806,
+ "grad_norm": 104.4194564819336,
+ "learning_rate": 1.5923172242874846e-05,
+ "loss": 6.1,
+ "step": 2810
+ },
+ {
+ "epoch": 10.481927710843374,
+ "grad_norm": 118.46540069580078,
+ "learning_rate": 1.5861214374225527e-05,
+ "loss": 7.1027,
+ "step": 2820
+ },
+ {
+ "epoch": 10.51899907321594,
+ "grad_norm": 108.23471069335938,
+ "learning_rate": 1.5799256505576208e-05,
+ "loss": 8.639,
+ "step": 2830
+ },
+ {
+ "epoch": 10.556070435588508,
+ "grad_norm": 144.46780395507812,
+ "learning_rate": 1.573729863692689e-05,
+ "loss": 6.9937,
+ "step": 2840
+ },
+ {
+ "epoch": 10.593141797961074,
+ "grad_norm": 108.4492416381836,
+ "learning_rate": 1.567534076827757e-05,
+ "loss": 7.2734,
+ "step": 2850
+ },
+ {
+ "epoch": 10.593141797961074,
+ "eval_loss": 2.8532466888427734,
+ "eval_runtime": 34.3677,
+ "eval_samples_per_second": 47.079,
+ "eval_steps_per_second": 6.751,
+ "step": 2850
+ },
+ {
+ "epoch": 10.630213160333643,
+ "grad_norm": 143.1586151123047,
+ "learning_rate": 1.5613382899628255e-05,
+ "loss": 7.6321,
+ "step": 2860
+ },
+ {
+ "epoch": 10.66728452270621,
+ "grad_norm": 98.63433837890625,
+ "learning_rate": 1.5551425030978936e-05,
+ "loss": 7.5788,
+ "step": 2870
+ },
+ {
+ "epoch": 10.704355885078776,
+ "grad_norm": 90.4648666381836,
+ "learning_rate": 1.5489467162329617e-05,
+ "loss": 6.7864,
+ "step": 2880
+ },
+ {
+ "epoch": 10.741427247451345,
+ "grad_norm": 141.69007873535156,
+ "learning_rate": 1.5427509293680298e-05,
+ "loss": 7.4237,
+ "step": 2890
+ },
+ {
+ "epoch": 10.778498609823911,
+ "grad_norm": 126.74280548095703,
+ "learning_rate": 1.536555142503098e-05,
+ "loss": 6.9813,
+ "step": 2900
+ },
+ {
+ "epoch": 10.815569972196478,
+ "grad_norm": 82.08953094482422,
+ "learning_rate": 1.5303593556381663e-05,
+ "loss": 6.6884,
+ "step": 2910
+ },
+ {
+ "epoch": 10.852641334569045,
+ "grad_norm": 123.8572998046875,
+ "learning_rate": 1.5241635687732344e-05,
+ "loss": 6.7464,
+ "step": 2920
+ },
+ {
+ "epoch": 10.889712696941613,
+ "grad_norm": 132.33477783203125,
+ "learning_rate": 1.5179677819083025e-05,
+ "loss": 7.7989,
+ "step": 2930
+ },
+ {
+ "epoch": 10.92678405931418,
+ "grad_norm": 81.7636489868164,
+ "learning_rate": 1.5117719950433704e-05,
+ "loss": 7.3568,
+ "step": 2940
+ },
+ {
+ "epoch": 10.963855421686747,
+ "grad_norm": 244.56277465820312,
+ "learning_rate": 1.5055762081784385e-05,
+ "loss": 8.6706,
+ "step": 2950
+ },
+ {
+ "epoch": 11.0,
+ "grad_norm": 121.42695617675781,
+ "learning_rate": 1.4993804213135068e-05,
+ "loss": 6.5687,
+ "step": 2960
+ },
+ {
+ "epoch": 11.037071362372567,
+ "grad_norm": 145.1683349609375,
+ "learning_rate": 1.493184634448575e-05,
+ "loss": 5.8992,
+ "step": 2970
+ },
+ {
+ "epoch": 11.074142724745135,
+ "grad_norm": 109.68660736083984,
+ "learning_rate": 1.4869888475836432e-05,
+ "loss": 6.4543,
+ "step": 2980
+ },
+ {
+ "epoch": 11.111214087117702,
+ "grad_norm": 93.12600708007812,
+ "learning_rate": 1.4807930607187113e-05,
+ "loss": 6.1386,
+ "step": 2990
+ },
+ {
+ "epoch": 11.148285449490269,
+ "grad_norm": 167.23182678222656,
+ "learning_rate": 1.4745972738537794e-05,
+ "loss": 6.9047,
+ "step": 3000
+ },
+ {
+ "epoch": 11.148285449490269,
+ "eval_loss": 2.9146969318389893,
+ "eval_runtime": 34.386,
+ "eval_samples_per_second": 47.054,
+ "eval_steps_per_second": 6.747,
+ "step": 3000
+ },
+ {
+ "epoch": 11.185356811862835,
+ "grad_norm": 115.30802154541016,
+ "learning_rate": 1.4684014869888477e-05,
+ "loss": 7.405,
+ "step": 3010
+ },
+ {
+ "epoch": 11.222428174235404,
+ "grad_norm": 103.70664978027344,
+ "learning_rate": 1.4622057001239158e-05,
+ "loss": 7.5441,
+ "step": 3020
+ },
+ {
+ "epoch": 11.25949953660797,
+ "grad_norm": 102.01170349121094,
+ "learning_rate": 1.4560099132589839e-05,
+ "loss": 6.7524,
+ "step": 3030
+ },
+ {
+ "epoch": 11.296570898980537,
+ "grad_norm": 151.57167053222656,
+ "learning_rate": 1.4498141263940521e-05,
+ "loss": 7.698,
+ "step": 3040
+ },
+ {
+ "epoch": 11.333642261353106,
+ "grad_norm": 158.46485900878906,
+ "learning_rate": 1.4436183395291202e-05,
+ "loss": 7.6167,
+ "step": 3050
+ },
+ {
+ "epoch": 11.370713623725672,
+ "grad_norm": 160.52752685546875,
+ "learning_rate": 1.4374225526641885e-05,
+ "loss": 7.1516,
+ "step": 3060
+ },
+ {
+ "epoch": 11.407784986098239,
+ "grad_norm": 120.98091888427734,
+ "learning_rate": 1.4312267657992566e-05,
+ "loss": 6.7458,
+ "step": 3070
+ },
+ {
+ "epoch": 11.444856348470806,
+ "grad_norm": 94.3011474609375,
+ "learning_rate": 1.4250309789343247e-05,
+ "loss": 6.7608,
+ "step": 3080
+ },
+ {
+ "epoch": 11.481927710843374,
+ "grad_norm": 95.0993881225586,
+ "learning_rate": 1.4188351920693928e-05,
+ "loss": 7.1508,
+ "step": 3090
+ },
+ {
+ "epoch": 11.51899907321594,
+ "grad_norm": 114.52859497070312,
+ "learning_rate": 1.4126394052044609e-05,
+ "loss": 6.9155,
+ "step": 3100
+ },
+ {
+ "epoch": 11.556070435588508,
+ "grad_norm": 124.27045440673828,
+ "learning_rate": 1.4064436183395292e-05,
+ "loss": 6.6664,
+ "step": 3110
+ },
+ {
+ "epoch": 11.593141797961074,
+ "grad_norm": 149.13070678710938,
+ "learning_rate": 1.4002478314745973e-05,
+ "loss": 8.3841,
+ "step": 3120
+ },
+ {
+ "epoch": 11.630213160333643,
+ "grad_norm": 93.60704040527344,
+ "learning_rate": 1.3940520446096655e-05,
+ "loss": 7.1934,
+ "step": 3130
+ },
+ {
+ "epoch": 11.66728452270621,
+ "grad_norm": 104.77314758300781,
+ "learning_rate": 1.3878562577447336e-05,
+ "loss": 6.9681,
+ "step": 3140
+ },
+ {
+ "epoch": 11.704355885078776,
+ "grad_norm": 150.2197265625,
+ "learning_rate": 1.3816604708798017e-05,
+ "loss": 7.2187,
+ "step": 3150
+ },
+ {
+ "epoch": 11.704355885078776,
+ "eval_loss": 2.7509171962738037,
+ "eval_runtime": 34.3139,
+ "eval_samples_per_second": 47.153,
+ "eval_steps_per_second": 6.761,
+ "step": 3150
+ },
+ {
+ "epoch": 11.741427247451345,
+ "grad_norm": 164.05996704101562,
+ "learning_rate": 1.37546468401487e-05,
+ "loss": 7.3155,
+ "step": 3160
+ },
+ {
+ "epoch": 11.778498609823911,
+ "grad_norm": 104.55535125732422,
+ "learning_rate": 1.3692688971499381e-05,
+ "loss": 7.3103,
+ "step": 3170
+ },
+ {
+ "epoch": 11.815569972196478,
+ "grad_norm": 100.98657989501953,
+ "learning_rate": 1.3630731102850064e-05,
+ "loss": 7.1959,
+ "step": 3180
+ },
+ {
+ "epoch": 11.852641334569045,
+ "grad_norm": 109.6310043334961,
+ "learning_rate": 1.3568773234200743e-05,
+ "loss": 6.8164,
+ "step": 3190
+ },
+ {
+ "epoch": 11.889712696941613,
+ "grad_norm": 116.8626480102539,
+ "learning_rate": 1.3506815365551424e-05,
+ "loss": 7.5836,
+ "step": 3200
+ },
+ {
+ "epoch": 11.92678405931418,
+ "grad_norm": 106.15950775146484,
+ "learning_rate": 1.3444857496902107e-05,
+ "loss": 5.2671,
+ "step": 3210
+ },
+ {
+ "epoch": 11.963855421686747,
+ "grad_norm": 131.19873046875,
+ "learning_rate": 1.3382899628252788e-05,
+ "loss": 6.4929,
+ "step": 3220
+ },
+ {
+ "epoch": 12.0,
+ "grad_norm": 61.416534423828125,
+ "learning_rate": 1.332094175960347e-05,
+ "loss": 7.0892,
+ "step": 3230
+ },
+ {
+ "epoch": 12.037071362372567,
+ "grad_norm": 110.03706359863281,
+ "learning_rate": 1.3258983890954152e-05,
+ "loss": 7.0877,
+ "step": 3240
+ },
+ {
+ "epoch": 12.074142724745135,
+ "grad_norm": 68.34326171875,
+ "learning_rate": 1.3197026022304833e-05,
+ "loss": 5.8302,
+ "step": 3250
+ },
+ {
+ "epoch": 12.111214087117702,
+ "grad_norm": 136.8495635986328,
+ "learning_rate": 1.3135068153655515e-05,
+ "loss": 5.6145,
+ "step": 3260
+ },
+ {
+ "epoch": 12.148285449490269,
+ "grad_norm": 103.39452362060547,
+ "learning_rate": 1.3073110285006196e-05,
+ "loss": 6.5808,
+ "step": 3270
+ },
+ {
+ "epoch": 12.185356811862835,
+ "grad_norm": 107.7593994140625,
+ "learning_rate": 1.3011152416356879e-05,
+ "loss": 6.6826,
+ "step": 3280
+ },
+ {
+ "epoch": 12.222428174235404,
+ "grad_norm": 99.79641723632812,
+ "learning_rate": 1.2949194547707558e-05,
+ "loss": 5.9819,
+ "step": 3290
+ },
+ {
+ "epoch": 12.25949953660797,
+ "grad_norm": 112.78658294677734,
+ "learning_rate": 1.2887236679058241e-05,
+ "loss": 6.68,
+ "step": 3300
+ },
+ {
+ "epoch": 12.25949953660797,
+ "eval_loss": 3.0174973011016846,
+ "eval_runtime": 34.2754,
+ "eval_samples_per_second": 47.206,
+ "eval_steps_per_second": 6.769,
+ "step": 3300
+ },
+ {
+ "epoch": 12.296570898980537,
+ "grad_norm": 139.285400390625,
+ "learning_rate": 1.2825278810408922e-05,
+ "loss": 6.1685,
+ "step": 3310
+ },
+ {
+ "epoch": 12.333642261353106,
+ "grad_norm": 121.69828796386719,
+ "learning_rate": 1.2763320941759603e-05,
+ "loss": 6.4473,
+ "step": 3320
+ },
+ {
+ "epoch": 12.370713623725672,
+ "grad_norm": 146.68112182617188,
+ "learning_rate": 1.2701363073110286e-05,
+ "loss": 6.3965,
+ "step": 3330
+ },
+ {
+ "epoch": 12.407784986098239,
+ "grad_norm": 144.31346130371094,
+ "learning_rate": 1.2639405204460967e-05,
+ "loss": 6.6278,
+ "step": 3340
+ },
+ {
+ "epoch": 12.444856348470806,
+ "grad_norm": 113.88147735595703,
+ "learning_rate": 1.257744733581165e-05,
+ "loss": 5.4575,
+ "step": 3350
+ },
+ {
+ "epoch": 12.481927710843374,
+ "grad_norm": 199.89317321777344,
+ "learning_rate": 1.251548946716233e-05,
+ "loss": 7.3019,
+ "step": 3360
+ },
+ {
+ "epoch": 12.51899907321594,
+ "grad_norm": 142.75885009765625,
+ "learning_rate": 1.2453531598513011e-05,
+ "loss": 7.4843,
+ "step": 3370
+ },
+ {
+ "epoch": 12.556070435588508,
+ "grad_norm": 185.7791748046875,
+ "learning_rate": 1.2391573729863692e-05,
+ "loss": 6.709,
+ "step": 3380
+ },
+ {
+ "epoch": 12.593141797961074,
+ "grad_norm": 100.56149291992188,
+ "learning_rate": 1.2329615861214373e-05,
+ "loss": 6.7168,
+ "step": 3390
+ },
+ {
+ "epoch": 12.630213160333643,
+ "grad_norm": 169.69461059570312,
+ "learning_rate": 1.2267657992565056e-05,
+ "loss": 7.0223,
+ "step": 3400
+ },
+ {
+ "epoch": 12.66728452270621,
+ "grad_norm": 175.87796020507812,
+ "learning_rate": 1.2205700123915737e-05,
+ "loss": 6.5089,
+ "step": 3410
+ },
+ {
+ "epoch": 12.704355885078776,
+ "grad_norm": 143.18231201171875,
+ "learning_rate": 1.214374225526642e-05,
+ "loss": 6.5094,
+ "step": 3420
+ },
+ {
+ "epoch": 12.741427247451345,
+ "grad_norm": 139.9623260498047,
+ "learning_rate": 1.20817843866171e-05,
+ "loss": 7.2317,
+ "step": 3430
+ },
+ {
+ "epoch": 12.778498609823911,
+ "grad_norm": 162.18252563476562,
+ "learning_rate": 1.2019826517967782e-05,
+ "loss": 6.6885,
+ "step": 3440
+ },
+ {
+ "epoch": 12.815569972196478,
+ "grad_norm": 146.02406311035156,
+ "learning_rate": 1.1957868649318464e-05,
+ "loss": 6.9693,
+ "step": 3450
+ },
+ {
+ "epoch": 12.815569972196478,
+ "eval_loss": 2.8461835384368896,
+ "eval_runtime": 34.1657,
+ "eval_samples_per_second": 47.357,
+ "eval_steps_per_second": 6.79,
+ "step": 3450
+ },
+ {
+ "epoch": 12.852641334569045,
+ "grad_norm": 124.50180053710938,
+ "learning_rate": 1.1895910780669145e-05,
+ "loss": 6.8242,
+ "step": 3460
+ },
+ {
+ "epoch": 12.889712696941613,
+ "grad_norm": 112.58306121826172,
+ "learning_rate": 1.1833952912019828e-05,
+ "loss": 6.6899,
+ "step": 3470
+ },
+ {
+ "epoch": 12.92678405931418,
+ "grad_norm": 122.80091857910156,
+ "learning_rate": 1.1771995043370507e-05,
+ "loss": 6.9113,
+ "step": 3480
+ },
+ {
+ "epoch": 12.963855421686747,
+ "grad_norm": 100.88201141357422,
+ "learning_rate": 1.1710037174721188e-05,
+ "loss": 7.1903,
+ "step": 3490
+ },
+ {
+ "epoch": 13.0,
+ "grad_norm": 91.95454406738281,
+ "learning_rate": 1.1648079306071871e-05,
+ "loss": 7.3286,
+ "step": 3500
+ },
+ {
+ "epoch": 13.037071362372567,
+ "grad_norm": 150.1581573486328,
+ "learning_rate": 1.1586121437422552e-05,
+ "loss": 6.5465,
+ "step": 3510
+ },
+ {
+ "epoch": 13.074142724745135,
+ "grad_norm": 124.90875244140625,
+ "learning_rate": 1.1524163568773235e-05,
+ "loss": 5.6804,
+ "step": 3520
+ },
+ {
+ "epoch": 13.111214087117702,
+ "grad_norm": 89.16959381103516,
+ "learning_rate": 1.1462205700123916e-05,
+ "loss": 5.6412,
+ "step": 3530
+ },
+ {
+ "epoch": 13.148285449490269,
+ "grad_norm": 139.5684356689453,
+ "learning_rate": 1.1400247831474599e-05,
+ "loss": 6.6161,
+ "step": 3540
+ },
+ {
+ "epoch": 13.185356811862835,
+ "grad_norm": 123.66468048095703,
+ "learning_rate": 1.133828996282528e-05,
+ "loss": 5.761,
+ "step": 3550
+ },
+ {
+ "epoch": 13.222428174235404,
+ "grad_norm": 92.05863189697266,
+ "learning_rate": 1.127633209417596e-05,
+ "loss": 5.5669,
+ "step": 3560
+ },
+ {
+ "epoch": 13.25949953660797,
+ "grad_norm": 106.46834564208984,
+ "learning_rate": 1.1214374225526643e-05,
+ "loss": 5.6184,
+ "step": 3570
+ },
+ {
+ "epoch": 13.296570898980537,
+ "grad_norm": 122.37000274658203,
+ "learning_rate": 1.1152416356877323e-05,
+ "loss": 6.2996,
+ "step": 3580
+ },
+ {
+ "epoch": 13.333642261353106,
+ "grad_norm": 115.14747619628906,
+ "learning_rate": 1.1090458488228005e-05,
+ "loss": 4.99,
+ "step": 3590
+ },
+ {
+ "epoch": 13.370713623725672,
+ "grad_norm": 123.00082397460938,
+ "learning_rate": 1.1028500619578686e-05,
+ "loss": 5.9974,
+ "step": 3600
+ },
+ {
+ "epoch": 13.370713623725672,
+ "eval_loss": 3.235833168029785,
+ "eval_runtime": 34.1503,
+ "eval_samples_per_second": 47.379,
+ "eval_steps_per_second": 6.793,
+ "step": 3600
+ },
+ {
+ "epoch": 13.407784986098239,
+ "grad_norm": 97.1900634765625,
+ "learning_rate": 1.0966542750929367e-05,
+ "loss": 5.6962,
+ "step": 3610
+ },
+ {
+ "epoch": 13.444856348470806,
+ "grad_norm": 166.885986328125,
+ "learning_rate": 1.090458488228005e-05,
+ "loss": 6.3662,
+ "step": 3620
+ },
+ {
+ "epoch": 13.481927710843374,
+ "grad_norm": 153.35101318359375,
+ "learning_rate": 1.0842627013630731e-05,
+ "loss": 7.0398,
+ "step": 3630
+ },
+ {
+ "epoch": 13.51899907321594,
+ "grad_norm": 159.07012939453125,
+ "learning_rate": 1.0780669144981414e-05,
+ "loss": 7.7358,
+ "step": 3640
+ },
+ {
+ "epoch": 13.556070435588508,
+ "grad_norm": 184.1981201171875,
+ "learning_rate": 1.0718711276332095e-05,
+ "loss": 7.9063,
+ "step": 3650
+ },
+ {
+ "epoch": 13.593141797961074,
+ "grad_norm": 125.76744842529297,
+ "learning_rate": 1.0656753407682777e-05,
+ "loss": 5.7823,
+ "step": 3660
+ },
+ {
+ "epoch": 13.630213160333643,
+ "grad_norm": 168.76263427734375,
+ "learning_rate": 1.0594795539033458e-05,
+ "loss": 6.9861,
+ "step": 3670
+ },
+ {
+ "epoch": 13.66728452270621,
+ "grad_norm": 195.7327423095703,
+ "learning_rate": 1.0532837670384138e-05,
+ "loss": 7.2855,
+ "step": 3680
+ },
+ {
+ "epoch": 13.704355885078776,
+ "grad_norm": 140.83375549316406,
+ "learning_rate": 1.047087980173482e-05,
+ "loss": 5.6785,
+ "step": 3690
+ },
+ {
+ "epoch": 13.741427247451345,
+ "grad_norm": 123.63497161865234,
+ "learning_rate": 1.0408921933085501e-05,
+ "loss": 6.4071,
+ "step": 3700
+ },
+ {
+ "epoch": 13.778498609823911,
+ "grad_norm": 80.74890899658203,
+ "learning_rate": 1.0346964064436184e-05,
+ "loss": 6.4294,
+ "step": 3710
+ },
+ {
+ "epoch": 13.815569972196478,
+ "grad_norm": 120.82400512695312,
+ "learning_rate": 1.0285006195786865e-05,
+ "loss": 6.0842,
+ "step": 3720
+ },
+ {
+ "epoch": 13.852641334569045,
+ "grad_norm": 152.11407470703125,
+ "learning_rate": 1.0223048327137546e-05,
+ "loss": 5.9422,
+ "step": 3730
+ },
+ {
+ "epoch": 13.889712696941613,
+ "grad_norm": 112.44628143310547,
+ "learning_rate": 1.0161090458488229e-05,
+ "loss": 7.0778,
+ "step": 3740
+ },
+ {
+ "epoch": 13.92678405931418,
+ "grad_norm": 210.1138916015625,
+ "learning_rate": 1.009913258983891e-05,
+ "loss": 8.1597,
+ "step": 3750
+ },
+ {
+ "epoch": 13.92678405931418,
+ "eval_loss": 3.0093190670013428,
+ "eval_runtime": 34.173,
+ "eval_samples_per_second": 47.347,
+ "eval_steps_per_second": 6.789,
+ "step": 3750
+ },
+ {
+ "epoch": 13.963855421686747,
+ "grad_norm": 157.09022521972656,
+ "learning_rate": 1.0037174721189593e-05,
+ "loss": 6.3154,
+ "step": 3760
+ },
+ {
+ "epoch": 14.0,
+ "grad_norm": 149.8700408935547,
+ "learning_rate": 9.975216852540272e-06,
+ "loss": 6.2416,
+ "step": 3770
+ },
+ {
+ "epoch": 14.037071362372567,
+ "grad_norm": 95.38166809082031,
+ "learning_rate": 9.913258983890955e-06,
+ "loss": 5.9958,
+ "step": 3780
+ },
+ {
+ "epoch": 14.074142724745135,
+ "grad_norm": 112.23654174804688,
+ "learning_rate": 9.851301115241636e-06,
+ "loss": 5.7032,
+ "step": 3790
+ },
+ {
+ "epoch": 14.111214087117702,
+ "grad_norm": 74.8873062133789,
+ "learning_rate": 9.789343246592317e-06,
+ "loss": 4.9524,
+ "step": 3800
+ },
+ {
+ "epoch": 14.148285449490269,
+ "grad_norm": 105.66299438476562,
+ "learning_rate": 9.727385377943e-06,
+ "loss": 5.386,
+ "step": 3810
+ },
+ {
+ "epoch": 14.185356811862835,
+ "grad_norm": 108.44865417480469,
+ "learning_rate": 9.66542750929368e-06,
+ "loss": 5.6353,
+ "step": 3820
+ },
+ {
+ "epoch": 14.222428174235404,
+ "grad_norm": 93.0536117553711,
+ "learning_rate": 9.603469640644363e-06,
+ "loss": 5.0873,
+ "step": 3830
+ },
+ {
+ "epoch": 14.25949953660797,
+ "grad_norm": 98.93279266357422,
+ "learning_rate": 9.541511771995044e-06,
+ "loss": 4.9255,
+ "step": 3840
+ },
+ {
+ "epoch": 14.296570898980537,
+ "grad_norm": 143.44992065429688,
+ "learning_rate": 9.479553903345725e-06,
+ "loss": 5.1423,
+ "step": 3850
+ },
+ {
+ "epoch": 14.333642261353106,
+ "grad_norm": 103.24166107177734,
+ "learning_rate": 9.417596034696408e-06,
+ "loss": 6.0775,
+ "step": 3860
+ },
+ {
+ "epoch": 14.370713623725672,
+ "grad_norm": 208.1169891357422,
+ "learning_rate": 9.355638166047087e-06,
+ "loss": 4.5073,
+ "step": 3870
+ },
+ {
+ "epoch": 14.407784986098239,
+ "grad_norm": 186.23892211914062,
+ "learning_rate": 9.29368029739777e-06,
+ "loss": 6.8347,
+ "step": 3880
+ },
+ {
+ "epoch": 14.444856348470806,
+ "grad_norm": 190.22607421875,
+ "learning_rate": 9.23172242874845e-06,
+ "loss": 6.5397,
+ "step": 3890
+ },
+ {
+ "epoch": 14.481927710843374,
+ "grad_norm": 186.79270935058594,
+ "learning_rate": 9.169764560099133e-06,
+ "loss": 7.2143,
+ "step": 3900
+ },
+ {
+ "epoch": 14.481927710843374,
+ "eval_loss": 3.308037042617798,
+ "eval_runtime": 41.7479,
+ "eval_samples_per_second": 38.756,
+ "eval_steps_per_second": 5.557,
+ "step": 3900
+ },
+ {
+ "epoch": 14.51899907321594,
+ "grad_norm": 146.21145629882812,
+ "learning_rate": 9.107806691449814e-06,
+ "loss": 6.1123,
+ "step": 3910
+ },
+ {
+ "epoch": 14.556070435588508,
+ "grad_norm": 170.6797637939453,
+ "learning_rate": 9.045848822800495e-06,
+ "loss": 6.6048,
+ "step": 3920
+ },
+ {
+ "epoch": 14.593141797961074,
+ "grad_norm": 118.28248596191406,
+ "learning_rate": 8.983890954151178e-06,
+ "loss": 6.3464,
+ "step": 3930
+ },
+ {
+ "epoch": 14.630213160333643,
+ "grad_norm": 170.6762237548828,
+ "learning_rate": 8.921933085501859e-06,
+ "loss": 6.3618,
+ "step": 3940
+ },
+ {
+ "epoch": 14.66728452270621,
+ "grad_norm": 148.5935516357422,
+ "learning_rate": 8.859975216852542e-06,
+ "loss": 6.5718,
+ "step": 3950
+ },
+ {
+ "epoch": 14.704355885078776,
+ "grad_norm": 72.61006164550781,
+ "learning_rate": 8.798017348203223e-06,
+ "loss": 5.9785,
+ "step": 3960
+ },
+ {
+ "epoch": 14.741427247451345,
+ "grad_norm": 164.15184020996094,
+ "learning_rate": 8.736059479553902e-06,
+ "loss": 6.5758,
+ "step": 3970
+ },
+ {
+ "epoch": 14.778498609823911,
+ "grad_norm": 140.56019592285156,
+ "learning_rate": 8.674101610904585e-06,
+ "loss": 6.4308,
+ "step": 3980
+ },
+ {
+ "epoch": 14.815569972196478,
+ "grad_norm": 127.03153991699219,
+ "learning_rate": 8.612143742255266e-06,
+ "loss": 6.0208,
+ "step": 3990
+ },
+ {
+ "epoch": 14.852641334569045,
+ "grad_norm": 146.550048828125,
+ "learning_rate": 8.550185873605949e-06,
+ "loss": 6.0303,
+ "step": 4000
+ },
+ {
+ "epoch": 14.889712696941613,
+ "grad_norm": 106.98975372314453,
+ "learning_rate": 8.48822800495663e-06,
+ "loss": 6.6396,
+ "step": 4010
+ },
+ {
+ "epoch": 14.92678405931418,
+ "grad_norm": 124.45453643798828,
+ "learning_rate": 8.42627013630731e-06,
+ "loss": 6.0184,
+ "step": 4020
+ },
+ {
+ "epoch": 14.963855421686747,
+ "grad_norm": 185.95437622070312,
+ "learning_rate": 8.364312267657993e-06,
+ "loss": 6.6248,
+ "step": 4030
+ },
+ {
+ "epoch": 15.0,
+ "grad_norm": 129.69049072265625,
+ "learning_rate": 8.302354399008674e-06,
+ "loss": 6.4538,
+ "step": 4040
+ },
+ {
+ "epoch": 15.037071362372567,
+ "grad_norm": 164.9100799560547,
+ "learning_rate": 8.240396530359357e-06,
+ "loss": 6.4742,
+ "step": 4050
+ },
+ {
+ "epoch": 15.037071362372567,
+ "eval_loss": 3.176128625869751,
+ "eval_runtime": 34.1148,
+ "eval_samples_per_second": 47.428,
+ "eval_steps_per_second": 6.801,
+ "step": 4050
+ }
+ ],
+ "logging_steps": 10,
+ "max_steps": 5380,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 20,
+ "save_steps": 150,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 0.0,
+ "train_batch_size": 7,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-4050/training_args.bin b/checkpoints/checkpoint-4050/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..f1a69c7344d84f80932397ffdc7afef25687d5b4
--- /dev/null
+++ b/checkpoints/checkpoint-4050/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fda94467724e48ac975735648aa63a75558d48b6c993506e918a3476acf80697
+size 129
diff --git a/checkpoints/checkpoint-5380/model.safetensors b/checkpoints/checkpoint-5380/model.safetensors
index 8f656a0b1796d8d04dedb7898b8a58d232f7eb37..96654dcaddd6cad04e0c3ce9121c5bd2432fff63 100644
--- a/checkpoints/checkpoint-5380/model.safetensors
+++ b/checkpoints/checkpoint-5380/model.safetensors
@@ -1,3 +1,3 @@
version https://git-lfs.github.com/spec/v1
-oid sha256:0aca6f2662c641a9490874428fd31456882c4bb2907af32f94152fb87b7038c1
-size 2239607176
+oid sha256:57286565ea2885d17737b0f38e37f3f918be92865015ac954c5b7c5344b9c6a0
+size 135
diff --git a/checkpoints/checkpoint-5380/optimizer.pt b/checkpoints/checkpoint-5380/optimizer.pt
index 08dbd12c42421caf14c56cf5a20a059f9eb74eb1..282d49167e7990774db93c5359863664a6aba05c 100644
--- a/checkpoints/checkpoint-5380/optimizer.pt
+++ b/checkpoints/checkpoint-5380/optimizer.pt
@@ -1,3 +1,3 @@
version https://git-lfs.github.com/spec/v1
-oid sha256:0f349ba6cca94122d3697d7432f2004a5b86a8df30ce3bf8064f9daf9dce426d
-size 4471055801
+oid sha256:06f8fcb363879a760902824000ea5cf1052dd93a7919440a3c13f2b237d778d4
+size 135
diff --git a/checkpoints/checkpoint-5380/rng_state.pth b/checkpoints/checkpoint-5380/rng_state.pth
index 46441ad0e51fbd6135b9c0414d044e1d405598e2..887f571a1286210e204d268c44474b8b4837c931 100644
--- a/checkpoints/checkpoint-5380/rng_state.pth
+++ b/checkpoints/checkpoint-5380/rng_state.pth
@@ -1,3 +1,3 @@
version https://git-lfs.github.com/spec/v1
-oid sha256:fed4ec41f6f10b5568c8f0515585b79470392fc10d7e8e323298c9286ef70ff4
-size 14244
+oid sha256:cc3c4ebadb2538040a72b967f07702fa7ff558e245f1f28942a1410a23fb5368
+size 130
diff --git a/checkpoints/checkpoint-5380/scheduler.pt b/checkpoints/checkpoint-5380/scheduler.pt
index e3ad878bba099afb7be5063b0df7c681376a4fa3..4c3f37111374adf0ed31989fbe2b78d254d0507e 100644
--- a/checkpoints/checkpoint-5380/scheduler.pt
+++ b/checkpoints/checkpoint-5380/scheduler.pt
@@ -1,3 +1,3 @@
version https://git-lfs.github.com/spec/v1
-oid sha256:66273c24b65120916e914c34a57e8e6171eeedc07221cc1455ae136612f473b9
-size 1064
+oid sha256:8f2f124ba6160a197ff26fa55b6a175c64f5f202f0a59a23e6bc747e8a2cf81c
+size 129
diff --git a/checkpoints/checkpoint-5380/tokenizer.json b/checkpoints/checkpoint-5380/tokenizer.json
index 2a51933f1ccb3cf68d53b877cbfa24734ada642f..5ed19c669b24d8e44e6e4c98b8ef4bb83d8cab46 100644
--- a/checkpoints/checkpoint-5380/tokenizer.json
+++ b/checkpoints/checkpoint-5380/tokenizer.json
@@ -1,3 +1,3 @@
version https://git-lfs.github.com/spec/v1
-oid sha256:883b037111086fd4dfebbbc9b7cee11e1517b5e0c0514879478661440f137085
-size 17082987
+oid sha256:de6f09c3f9b891e5b98dd3af9463dcab5a97d5265e288271395324a0577e6c05
+size 133
diff --git a/checkpoints/checkpoint-5380/training_args.bin b/checkpoints/checkpoint-5380/training_args.bin
index 7d8a74599946218ee47c496ec84d07b167039f76..f1a69c7344d84f80932397ffdc7afef25687d5b4 100644
--- a/checkpoints/checkpoint-5380/training_args.bin
+++ b/checkpoints/checkpoint-5380/training_args.bin
@@ -1,3 +1,3 @@
version https://git-lfs.github.com/spec/v1
-oid sha256:1cdcceaae6653fb2ddf52aed92e50c566c84ffbec2efeda566b6d2719f722937
-size 5624
+oid sha256:fda94467724e48ac975735648aa63a75558d48b6c993506e918a3476acf80697
+size 129