Upload folder using huggingface_hub

.gitattributes

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

README.md

@@ -0,0 +1,127 @@
+---
+license: apache-2.0
+base_model: Qwen/Qwen2.5-Coder-7B-Instruct
+pipeline_tag: text-generation
+library_name: transformers
+language:
+- en
+tags:
+- text-to-sql
+- sql
+- bird
+- chain-of-thought
+- reasoning
+- qwen2.5-coder
+- gguf
+- llama.cpp
+datasets:
+- jk200201/bird-cot-sft
+model-index:
+- name: qwen2.5-coder-7b-bird-cot
+  results:
+  - task:
+      type: text-generation
+      name: Text-to-SQL
+    dataset:
+      type: bird
+      name: BIRD (dev)
+    metrics:
+    - type: accuracy
+      name: Result accuracy (greedy)
+      value: 52.1
+    - type: accuracy
+      name: Result accuracy (self-consistency, K=8)
+      value: 58.5
+---
+
+# Qwen2.5-Coder-7B — BIRD CoT (Text-to-SQL)
+
+A 7B text-to-SQL model that **reasons step-by-step over a database schema, then writes the SQL**. Fine-tuned from `Qwen/Qwen2.5-Coder-7B-Instruct` by distilling *execution-verified* chain-of-thought solutions.
+
+On **BIRD dev** (messy, real-world schemas — the hard text-to-SQL benchmark) it reaches **52.1%** result accuracy greedy, and **58.5%** with self-consistency (Best-of-N, K=8) — **surpassing single-shot frontier models at a fraction of the size.**
+
+## Results — BIRD dev (execution result accuracy)
+
+| Model | Result accuracy |
+|---|---|
+| Qwen2.5-Coder-7B-Instruct (base) | ~27% |
+| **This model — greedy (1 sample)** | **52.1%** |
+| **This model — self-consistency (K=8)** | **58.5%** |
+| Grok-4 (single-shot) | 55.4% |
+| DeepSeek-V3 (single-shot) | 54.7% |
+
+*Result accuracy* = the generated SQL executes to the **same rows** as the gold query (BIRD's official execution metric). Frontier numbers are single-shot; the 58.5% uses K=8 self-consistency (8× inference).
+
+## Usage
+
+Prompt the model to reason step-by-step; it returns the reasoning followed by a fenced SQL block. **Take the last ```sql``` block** as the query.
+
+```python
+import re, torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+model_id = "jk200201/qwen2.5-coder-7b-bird-cot"
+tok = AutoTokenizer.from_pretrained(model_id)
+model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.bfloat16, device_map="auto")
+
+SYSTEM = ("You are an expert SQLite query writer. Reason step by step about the schema "
+          "and the question, then output the final query in a ```sql code block.")
+
+def build_prompt(schema, question):
+    return ("Given the database schema and question, work out the correct SQLite query step by step.\n\n"
+            f"Database Schema:\n{schema}\n\nQuestion: {question}\n\n"
+            "Think step by step:\n1. Which tables and columns are relevant?\n"
+            "2. What joins, filters, grouping, and ordering are needed?\n3. Handle edge cases.\n\n"
+            "Then give the final answer as:\n```sql\n<final query>\n```")
+
+schema   = "CREATE TABLE singer (Singer_ID INT, Name TEXT, Age INT);"
+question = "How many singers are older than 40?"
+messages = [{"role": "system", "content": SYSTEM},
+            {"role": "user",   "content": build_prompt(schema, question)}]
+
+text   = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+inputs = tok(text, return_tensors="pt").to(model.device)
+out    = model.generate(**inputs, max_new_tokens=512, do_sample=False)
+resp   = tok.decode(out[0][inputs.input_ids.shape[-1]:], skip_special_tokens=True)
+
+sql = re.findall(r"```(?:sql)?\s*(.*?)```", resp, re.DOTALL)[-1].strip()
+print(sql)
+```
+
+**For best accuracy (58.5%):** sample K=8 at `temperature=0.8`, execute each candidate, and take the majority result (self-consistency).
+
+## Run locally (GGUF / Ollama / llama.cpp)
+
+Quantized GGUF builds (`Q4_K_M`, `Q5_K_M`, `Q8_0`) are provided for CPU/laptop use:
+
+```bash
+ollama run jk200201/qwen2.5-coder-7b-bird-cot
+# or with llama.cpp:
+./llama-cli -m qwen2.5-coder-7b-bird-cot.Q4_K_M.gguf -sys "$SYSTEM" -p "$PROMPT"
+```
+
+## Training
+
+- **Method — reasoning distillation (CoT-SFT):** a strong teacher (Qwen3-Coder-480B) generated step-by-step CoT solutions on BIRD train; only **execution-verified-correct** chains were kept (**5,593** examples), then supervised-fine-tuned into the 7B. Distilling the teacher's *reasoning* generalized across BIRD's cross-domain dev databases better than distilling SQL answers directly.
+- **Config:** QLoRA (4-bit NF4, LoRA r=32, α=64), 2 epochs, LR 2e-4 cosine, max seq len 8192.
+- **Data:** [`jk200201/bird-cot-sft`](https://huggingface.co/datasets/jk200201/bird-cot-sft)
+
+## Limitations
+
+- Tuned for **BIRD-style** analytic SQL over realistic schemas; unusual dialects/domains may need adaptation. Emits **SQLite** dialect.
+- Greedy (52.1%) is the deployable single-shot; the 58.5% figure requires K=8 self-consistency (8× inference cost).
+- A 7B model — always review generated SQL before running it on production data.
+
+## Citation
+
+```bibtex
+@misc{qwen25coder7b_bird_cot_2026,
+  title  = {Qwen2.5-Coder-7B BIRD CoT: reasoning distillation for text-to-SQL},
+  author = {Jenish Kothari},
+  year   = {2026},
+  howpublished = {\url{https://huggingface.co/jk200201/qwen2.5-coder-7b-bird-cot}}
+}
+```
+
+## Acknowledgements
+Base: Qwen2.5-Coder (Alibaba Qwen). Teacher: Qwen3-Coder-480B (W&B Inference). Benchmark: [BIRD](https://bird-bench.github.io/).

chat_template.jinja

@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

config.json

@@ -0,0 +1,61 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pad_token_id": null,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 1000000.0,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "transformers_version": "5.13.1",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "5.13.1"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:432a36746704cd4f5bfb157a6e0884c74818de60bfe643cc606d847c246df57e
+size 15231272152

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892

tokenizer_config.json

@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": false,
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}