Update fine-tuned adapter weights

README.md

@@ -1,198 +1,202 @@
 ---
 base_model: Qwen/Qwen2.5-7B
 library_name: peft
-language:
-- en
-license: agpl-3.0
-datasets:
-- OramaSearch/nlp-to-query-small
 ---
 
-# Query Translator Mini
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
 
-This repository contains a fine-tuned version of Qwen 2.5 7B model specialized in translating natural language queries into structured Orama search queries.
 
-The model uses PEFT with LoRA to maintain efficiency while achieving high performance.
 
 ## Model Details
 
 ### Model Description
 
-The Query Translator Mini model is designed to convert natural language queries into structured JSON queries compatible with the Orama search engine.
-
-It understands various data types and query operators, making it versatile for different search scenarios.
-
-### Key Features
-
-- Translates natural language to structured Orama queries
-- Supports multiple field types: string, number, boolean, enum, and arrays
-- Handles complex query operators: `gt`, `gte`, `lt`, `lte`, `eq`, `between`, `containsAll`
-- Supports nested properties with dot notation
-- Works with both full-text search and filtered queries
-
-## Usage
-
-```python
-from transformers import AutoModelForCausalLM, AutoTokenizer
-from peft import PeftModel
-
-# Load the model and tokenizer
-model_name = "your-username/query-translator-mini"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForCausalLM.from_pretrained(model_name)
-
-# System Prompt used during training
-SYSTEM_PROMPT = """
-You are a tool used to generate synthetic data of Orama queries. Orama is a full-text, vector, and hybrid search engine.
-
-Let me show you what you need to do with some examples.
-
-Example:
-  - Query: `"What are the red wines that cost less than 20 dollars?"`
-  - Schema: `{ "name": "string", "content": "string", "price": "number", "tags": "enum[]" }`
-  - Generated query: `{ "term": "", "where": { "tags": { "containsAll": ["red", "wine"] }, "price": { "lt": 20 } } }`
-
-Another example:
-  - Query: `"Show me 5 prosecco wines good for aperitif"`
-  - Schema: `{ "name": "string", "content": "string", "price": "number", "tags": "enum[]" }`
-  - Generated query: `{ "term": "prosecco aperitif", "limit": 5 }`
-
-One last example:
-  - Query: `"Show me some wine reviews with a score greater than 4.5 and less than 5.0."`
-  - Schema: `{ "title": "string", "content": "string", "reviews": { "score": "number", "text": "string" } }]`
-  - Generated query: `{ "term": "", "where": { "reviews.score": { "between": [4.5, 5.0] } } }`
-
-The rules to generate the query are:
-
-- Never use an "embedding" field in the schema.
-- Every query has a "term" field that is a string. It represents the full-text search terms. Can be empty (will match all documents).
-- You can use a "where" field that is an object. It represents the filters to apply to the documents. Its keys and values depend on the schema of the database:
-  - If the field is a "string", you should not use operators. Example: `{ "where": { "title": "champagne" } }`.
-  - If the field is a "number", you can use the following operators: "gt", "gte", "lt", "lte", "eq", "between". Example: `{ "where": { "price": { "between": [20, 100] } } }`. Another example: `{ "where": { "price": { "lt": 20 } } }`.
-  - If the field is an "enum", you can use the following operators: "eq", "in", "nin". Example: `{ "where": { "tags": { "containsAll": ["red", "wine"] } } }`.
-  - If the field is an "string[]", it's gonna be just like the "string" field, but you can use an array of values. Example: `{ "where": { "title": ["champagne", "montagne"] } }`.
-  - If the field is a "boolean", you can use the following operators: "eq". Example: `{ "where": { "isAvailable": true } }`. Another example: `{ "where": { "isAvailable": false } }`.
-  - If the field is a "enum[]", you can use the following operators: "containsAll". Example: `{ "where": { "tags": { "containsAll": ["red", "wine"] } } }`.
-  - Nested properties are supported. Just translate them into dot notation. Example: `{ "where": { "author.name": "John" } }`.
-  - Array of numbers are not supported.
-  - Array of booleans are not supported.
-"""
-
-# Example query
-query = "What are the red wines that cost less than 20 dollars?"
-
-# Orama schema
-schema = {
-    "name": "string",
-    "content": "string",
-    "price": "number",
-    "tags": "enum[]"
-}
-
-# Generate structured query
-messages = [
-    {"role": "system", "content": SYSTEM_PROMPT},
-    {"role": "user", "content": f"Query: {query}\nSchema: {json.dumps(schema)}"},
-]
-
-prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
-inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
-outputs = model.generate(
-    **inputs,
-    max_length=512,
-    temperature=0.1,
-    top_p=0.9,
-    num_return_sequences=1,
-)
-
-response = tokenizer.decode(outputs[0], skip_special_tokens=True)
-```
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
 
 ## Training Details
 
-The model was trained using the following configuration:
-
-- Base Model: Qwen 2.5 7B
-- Training Method: LoRA
-- Quantization: 4-bit quantization using bitsandbytes
-- LoRA Configuration:
-  - Rank: 16
-  - Alpha: 32
-  - Dropout: 0.1
-  - Target Modules: Attention layers and MLP
-
-- Training Arguments:
-  - Epochs: 3
-  - Batch Size: 2
-  - Learning Rate: 5e-5
-  - Gradient Accumulation Steps: 8
-  - FP16 Training: Enabled
-  - Gradient Checkpointing: Enabled
-
-## Supported Query Types
-
-The model can handle various types of queries including:
-
-1. Simple text search:
-
-```json
-{
-    "term": "prosecco aperitif",
-    "limit": 5
-}
-```
-
-2. Numeric range queries:
-
-```json
-{
-    "term": "",
-    "where": {
-        "price": {
-            "between": [20, 100]
-        }
-    }
-}
-```
-
-3. Tag-based filtering:
-
-```json
-{
-    "term": "",
-    "where": {
-        "tags": {
-            "containsAll": ["red", "wine"]
-        }
-    }
-}
-```
-
-## Limitations
-
-- Does not support array of numbers or booleans
-- Maximum input length is 1024 tokens
-- Embedding fields are not supported in the schema
-
-## Citation
-
-If you use this model in your research, please cite:
-
-```
-@misc{query-translator-mini,
-  author = {OramaSearch Inc.},
-  title = {Query Translator Mini: Natural Language to Orama Query Translation},
-  year = {2024},
-  publisher = {HuggingFace},
-  journal = {HuggingFace Repository},
-  howpublished = {\url{https://huggingface.co/OramaSearch/query-translator-mini}}
-}
-```
-
-## License
-
-AGPLv3
-
-## Acknowledgments
-
-This model builds upon the Qwen 2.5 7B model and uses techniques from the PEFT library. Special thanks to the teams behind these projects.
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.14.0

adapter_config.json

@@ -17,19 +17,22 @@
   "lora_dropout": 0.1,
   "megatron_config": null,
   "megatron_core": "megatron.core",
-  "modules_to_save": null,
+  "modules_to_save": [
+    "embed_tokens",
+    "lm_head"
+  ],
   "peft_type": "LORA",
   "r": 16,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "self_attn.o_proj",
     "self_attn.v_proj",
-    "mlp.gate_proj",
+    "self_attn.q_proj",
+    "self_attn.k_proj",
     "mlp.up_proj",
+    "self_attn.o_proj",
     "mlp.down_proj",
-    "self_attn.k_proj",
-    "self_attn.q_proj"
+    "mlp.gate_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:03d12df6255da192aabbcb330241f396d88a6ccff51cd47049dba8ff9e0462bb
-size 161533192
+oid sha256:0fe5daa38880f7e6469caf0c77c4798fabcb7e9755229d1642c339a795c18942
+size 2260783880