Text Generation
PEFT
Safetensors
GGUF
Transformers
English
Spanish
harbour
fivewin
fwh
lora
sft
trl
unsloth
code-generation
xbase
clipper
conversational
Instructions to use fivetech/Harbour with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use fivetech/Harbour with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("/home/fivetech/finetune/models/Qwen3.6-35B-A3B") model = PeftModel.from_pretrained(base_model, "fivetech/Harbour") - Transformers
How to use fivetech/Harbour with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="fivetech/Harbour") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("fivetech/Harbour", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use fivetech/Harbour with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf fivetech/Harbour:Q4_K_M # Run inference directly in the terminal: llama cli -hf fivetech/Harbour:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf fivetech/Harbour:Q4_K_M # Run inference directly in the terminal: llama cli -hf fivetech/Harbour:Q4_K_M
Use pre-built binary
# Download pre-built binary from: # https://github.com/ggerganov/llama.cpp/releases # Start a local OpenAI-compatible server with a web UI: ./llama-server -hf fivetech/Harbour:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf fivetech/Harbour:Q4_K_M
Build from source code
git clone https://github.com/ggerganov/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-server llama-cli # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama-server -hf fivetech/Harbour:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf fivetech/Harbour:Q4_K_M
Use Docker
docker model run hf.co/fivetech/Harbour:Q4_K_M
- LM Studio
- Jan
- vLLM
How to use fivetech/Harbour with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "fivetech/Harbour" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "fivetech/Harbour", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/fivetech/Harbour:Q4_K_M
- SGLang
How to use fivetech/Harbour with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "fivetech/Harbour" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "fivetech/Harbour", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "fivetech/Harbour" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "fivetech/Harbour", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Ollama
How to use fivetech/Harbour with Ollama:
ollama run hf.co/fivetech/Harbour:Q4_K_M
- Unsloth Studio
How to use fivetech/Harbour with Unsloth Studio:
Install Unsloth Studio (macOS, Linux, WSL)
curl -fsSL https://unsloth.ai/install.sh | sh # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for fivetech/Harbour to start chatting
Install Unsloth Studio (Windows)
irm https://unsloth.ai/install.ps1 | iex # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for fivetech/Harbour to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for fivetech/Harbour to start chatting
- Pi
How to use fivetech/Harbour with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf fivetech/Harbour:Q4_K_M
Configure the model in Pi
# Install Pi: npm install -g @mariozechner/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "fivetech/Harbour:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent new
How to use fivetech/Harbour with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf fivetech/Harbour:Q4_K_M
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default fivetech/Harbour:Q4_K_M
Run Hermes
hermes
- Atomic Chat new
- OpenClaw new
How to use fivetech/Harbour with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf fivetech/Harbour:Q4_K_M
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "fivetech/Harbour:Q4_K_M" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
- Docker Model Runner
How to use fivetech/Harbour with Docker Model Runner:
docker model run hf.co/fivetech/Harbour:Q4_K_M
- Lemonade
How to use fivetech/Harbour with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull fivetech/Harbour:Q4_K_M
Run and chat with the model
lemonade run user.Harbour-Q4_K_M
List all available models
lemonade list
File size: 22,361 Bytes
0dfbe21 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 | #!/usr/bin/env python3
"""
Harbour Test Battery - Generates code, compiles with harbour, evaluates with qwen3.6:35b
"""
import json
import time
import subprocess
import requests
import tempfile
import os
from pathlib import Path
from datetime import datetime
OLLAMA_URL = "http://localhost:11434/api/generate"
MODEL = "qwen3.6:35b"
HARBOUR = "/home/fivetech/harbour/bin/linux/gcc/harbour"
WORK_DIR = Path("/home/fivetech/finetune/test_output")
WORK_DIR.mkdir(exist_ok=True)
def query_ollama(prompt, system="", timeout=300):
payload = {
"model": MODEL,
"prompt": prompt,
"stream": False,
"options": {"temperature": 0.2, "num_predict": 3000, "top_p": 0.9}
}
if system:
payload["system"] = system
try:
start = time.time()
r = requests.post(OLLAMA_URL, json=payload, timeout=timeout)
elapsed = time.time() - start
data = r.json()
return {
"response": data.get("response", ""),
"eval_count": data.get("eval_count", 0),
"duration": elapsed,
"tps": data.get("eval_count", 0) / max(data.get("eval_duration", 1) / 1e9, 0.001),
"error": None
}
except Exception as e:
return {"response": "", "error": str(e), "eval_count": 0, "duration": 0, "tps": 0}
def compile_harbour(code):
"""Compile code with harbour, return (success, error_msg, obj_exists)"""
prg_file = WORK_DIR / "test.prg"
prg_file.write_text(code)
try:
result = subprocess.run(
[HARBOUR, str(prg_file), "-n", "-w"],
capture_output=True, text=True, timeout=30
)
obj_file = WORK_DIR / "test.obj"
success = result.returncode == 0
obj_exists = obj_file.exists()
error = result.stderr.strip() if result.stderr else ""
if not success and not error:
error = result.stdout.strip()
return success, error, obj_exists
except subprocess.TimeoutExpired:
return False, "Compilation timeout", False
except Exception as e:
return False, str(e), False
def clean_code(response):
"""Extract code from model response, remove markdown."""
lines = response.split('\n')
in_code = False
code_lines = []
skip_explanation = True
for line in lines:
stripped = line.strip()
# Skip markdown
if stripped.startswith('```'):
in_code = not in_code
continue
if in_code:
code_lines.append(line)
skip_explanation = False
elif skip_explanation:
# Detect start of code
upper = stripped.upper()
if any(upper.startswith(kw) for kw in [
'FUNCTION', 'PROCEDURE', 'LOCAL', 'STATIC', 'PUBLIC',
'PRIVATE', 'MEMVAR', '#DEFINE', '#INCLUDE', 'CLASS',
'METHOD', 'RETURN', 'SET', 'REQUEST'
]):
in_code = True
code_lines.append(line)
skip_explanation = False
if not code_lines:
# Fallback: take everything
code_lines = response.split('\n')
return '\n'.join(code_lines).strip()
# ============================================================
# TEST DEFINITIONS - Based on dataset patterns
# ============================================================
TESTS = [
# ---- BASIC SYNTAX ----
{
"id": "SYNTAX_01", "category": "Basic Syntax", "name": "Variable types and declarations",
"prompt": "Write a Harbour program that declares LOCAL variables of each type (numeric, character, logical, date, nil), prints them with ValType(), and uses proper Hungarian notation.",
"expected_keywords": ["LOCAL", "ValType", "FUNCTION"],
"min_lines": 8,
},
{
"id": "SYNTAX_02", "category": "Basic Syntax", "name": "Preprocessor defines",
"prompt": "Write Harbour preprocessor definitions for application constants: app name, version, max records, date format. Use #define and show conditional compilation with #ifdef.",
"expected_keywords": ["#DEFINE", "#IFDEF", "#ENDIF"],
"min_lines": 6,
},
{
"id": "SYNTAX_03", "category": "Basic Syntax", "name": "String operations",
"prompt": "Write a Harbour function that takes a full name string and returns initials. Use AllTrim, Upper, Left, At, SubStr, and Space functions.",
"expected_keywords": ["FUNCTION", "AllTrim", "Upper", "Left", "At", "SubStr"],
"min_lines": 6,
},
{
"id": "SYNTAX_04", "category": "Basic Syntax", "name": "Date functions",
"prompt": "Write a Harbour function that calculates the number of business days between two dates, excluding weekends. Use Date(), DOW(), and date arithmetic.",
"expected_keywords": ["FUNCTION", "Date", "DOW"],
"min_lines": 8,
},
{
"id": "SYNTAX_05", "category": "Basic Syntax", "name": "Type conversion",
"prompt": "Write Harbour code that converts between all types: Str, Val, CTOD, DTOC, ASC, Chr, Transform. Show edge cases.",
"expected_keywords": ["Str", "Val", "CTOD", "DTOC"],
"min_lines": 8,
},
# ---- CONTROL FLOW ----
{
"id": "CTRL_01", "category": "Control Flow", "name": "IF/ELSEIF/ENDIF",
"prompt": "Write a Harbour function that classifies employee salary into tax brackets using IF/ELSEIF/ELSE/ENDIF. Include 5 brackets and error handling.",
"expected_keywords": ["FUNCTION", "IF", "ELSEIF", "ELSE", "ENDIF"],
"min_lines": 10,
},
{
"id": "CTRL_02", "category": "Control Flow", "name": "DO CASE",
"prompt": "Write a Harbour function using DO CASE to convert month number (1-12) to season name. Handle invalid input with OTHERWISE.",
"expected_keywords": ["DO CASE", "CASE", "OTHERWISE", "ENDCASE"],
"min_lines": 8,
},
{
"id": "CTRL_03", "category": "Control Flow", "name": "FOR/NEXT loop",
"prompt": "Write a Harbour function using FOR/NEXT to calculate the sum of all prime numbers below 100. Include STEP and EXIT.",
"expected_keywords": ["FOR", "TO", "NEXT", "IF", "EXIT"],
"min_lines": 10,
},
{
"id": "CTRL_04", "category": "Control Flow", "name": "DO WHILE",
"prompt": "Write a Harbour function using DO WHILE to implement the Euclidean algorithm for GCD. Include LOOP and EXIT.",
"expected_keywords": ["DO WHILE", "ENDDO", "IF", "LOOP", "EXIT"],
"min_lines": 6,
},
{
"id": "CTRL_05", "category": "Control Flow", "name": "SCAN/ENDSCAN",
"prompt": "Write Harbour code using SCAN/ENDSCAN to find the longest string in an array. Include NEXT clause.",
"expected_keywords": ["SCAN", "ENDSCAN"],
"min_lines": 6,
},
{
"id": "CTRL_06", "category": "Control Flow", "name": "FOR EACH",
"prompt": "Write Harbour code using FOR EACH to count word frequencies in a string. Use a hash for storage.",
"expected_keywords": ["FOR EACH", "NEXT", ":="],
"min_lines": 8,
},
# ---- FUNCTIONS ----
{
"id": "FUNC_01", "category": "Functions", "name": "Parameters and return",
"prompt": "Write a Harbour function with default parameters, pass-by-reference using @, and return an array. Include proper Hungarian notation.",
"expected_keywords": ["FUNCTION", "LOCAL", "RETURN"],
"min_lines": 6,
},
{
"id": "FUNC_02", "category": "Functions", "name": "Recursion",
"prompt": "Write a recursive Harbour function for Fibonacci numbers with memoization using a hash. Include base case and error handling.",
"expected_keywords": ["FUNCTION", "IF", "RETURN"],
"min_lines": 8,
},
{
"id": "FUNC_03", "category": "Functions", "name": "Variable scope",
"prompt": "Write Harbour code demonstrating LOCAL, STATIC, PRIVATE, PUBLIC variables. Show scope differences with nested function calls.",
"expected_keywords": ["LOCAL", "STATIC", "PRIVATE", "PUBLIC"],
"min_lines": 8,
},
{
"id": "FUNC_04", "category": "Functions", "name": "Code blocks",
"prompt": "Write Harbour code using code blocks: AEval with {|x| x*2}, AScan, ASort with custom sort. Show evaluation with Eval().",
"expected_keywords": ["AEval", "AScan", "ASort", "Eval"],
"min_lines": 6,
},
{
"id": "FUNC_05", "category": "Functions", "name": "Error handling",
"prompt": "Write a Harbour function with BEGIN SEQUENCE/RECOVER/END SEQUENCE for file reading. Include DEFAULT and BREAK.",
"expected_keywords": ["BEGIN SEQUENCE", "RECOVER", "END SEQUENCE"],
"min_lines": 8,
},
# ---- ARRAYS ----
{
"id": "ARRAY_01", "category": "Arrays", "name": "Array operations",
"prompt": "Write Harbour functions for: create 2D array, AAdd elements, ASort with custom order, AScan by value, ASize to resize. Include error handling.",
"expected_keywords": ["ARRAY", "AAdd", "ASort", "AScan", "ASize"],
"min_lines": 8,
},
{
"id": "ARRAY_02", "category": "Arrays", "name": "Hash operations",
"prompt": "Write Harbour code using hashes: create, add keys, iterate with FOR EACH, merge two hashes, check key existence with HB_HHasKey, convert to array.",
"expected_keywords": [":=", "FOR EACH", "HB_HHasKey"],
"min_lines": 8,
},
{
"id": "ARRAY_03", "category": "Arrays", "name": "Sorting algorithm",
"prompt": "Implement QuickSort in Harbour for an array of numbers. Include partition logic and proper recursion.",
"expected_keywords": ["FUNCTION", "LOCAL", "IF", "RETURN"],
"min_lines": 12,
},
# ---- OOP ----
{
"id": "OOP_01", "category": "OOP", "name": "Class definition",
"prompt": "Write a Harbour class Person with DATA (name, age), METHOD (New constructor, GetName, SetAge), and CLASSDATA. Include validation in SetAge.",
"expected_keywords": ["CLASS", "DATA", "METHOD", "RETURN"],
"min_lines": 10,
},
{
"id": "OOP_02", "category": "OOP", "name": "Inheritance",
"prompt": "Write Harbour classes: Shape (base), Circle (derived) with area() method. Show inheritance syntax and method override.",
"expected_keywords": ["CLASS", "METHOD", "INHERIT"],
"min_lines": 10,
},
{
"id": "OOP_03", "category": "OOP", "name": "Operator overloading",
"prompt": "Write a Harbour class Vec2 for 2D vectors. Overload + and - operators. Include magnitude and normalize methods.",
"expected_keywords": ["CLASS", "METHOD", "OPERATOR"],
"min_lines": 12,
},
{
"id": "OOP_04", "category": "OOP", "name": "Singleton pattern",
"prompt": "Implement Singleton pattern in Harbour for a config manager. Ensure only one instance exists.",
"expected_keywords": ["CLASS", "CLASSDATA", "METHOD"],
"min_lines": 10,
},
# ---- DATABASE ----
{
"id": "DB_01", "category": "Database", "name": "Basic RDD",
"prompt": "Write Harbour code that creates a DBF file, opens it, appends records, and closes properly. Use DBCreate and DBUseArea.",
"expected_keywords": ["DBCreate", "DBUseArea", "DBAppend", "DBCLOSEALL"],
"min_lines": 10,
},
{
"id": "DB_02", "category": "Database", "name": "Indexing",
"prompt": "Write Harbour code creating an index on a DBF field using RDD. Include ORDSCOPE for range queries.",
"expected_keywords": ["ORDCREATE", "ORDSCOPE"],
"min_lines": 8,
},
{
"id": "DB_03", "category": "Database", "name": "DBEval",
"prompt": "Write Harbour code using DBEval to process all records: count, sum field values, and mark records meeting a condition.",
"expected_keywords": ["DBEval", "FOR", "WHILE"],
"min_lines": 8,
},
# ---- FILE I/O ----
{
"id": "FILE_01", "category": "File I/O", "name": "Text file read/write",
"prompt": "Write Harbour functions to read a text file line by line and write processed output. Use FCreate, FOpen, FRead, FWrite, FClose, FEof.",
"expected_keywords": ["FCreate", "FOpen", "FRead", "FWrite", "FClose", "FEof"],
"min_lines": 10,
},
{
"id": "FILE_02", "category": "File I/O", "name": "Directory listing",
"prompt": "Write Harbour code using Directory() to list files with a pattern, get file size and date, and process each file.",
"expected_keywords": ["Directory", "LEN", "FOR"],
"min_lines": 6,
},
# ---- COMPLEX ----
{
"id": "CMPX_01", "category": "Complex", "name": "CSV parser",
"prompt": "Write a Harbour CSV parser that reads a CSV file, handles quoted fields, and returns an array of arrays. Include error handling.",
"expected_keywords": ["FUNCTION", "LOCAL", "FClose", "FEof"],
"min_lines": 15,
},
{
"id": "CMPX_02", "category": "Complex", "name": "INI file reader",
"prompt": "Write a Harbour INI file parser. Read sections, keys, and values into a hash. Handle comments and empty lines.",
"expected_keywords": ["FUNCTION", "LOCAL", "HASH"],
"min_lines": 12,
},
{
"id": "CMPX_03", "category": "Complex", "name": "String template engine",
"prompt": "Write a Harbour template engine replacing {{variable}} placeholders with hash values. Include error handling for missing keys.",
"expected_keywords": ["FUNCTION", "LOCAL", "STRTRAN"],
"min_lines": 8,
},
{
"id": "CMPX_04", "category": "Complex", "name": "Logger",
"prompt": "Write a Harbour logging system with DEBUG/INFO/WARN/ERROR levels, timestamp, file output, and configurable level filtering.",
"expected_keywords": ["FUNCTION", "LOCAL", "FClose"],
"min_lines": 12,
},
{
"id": "CMPX_05", "category": "Complex", "name": "Base64 encoder",
"prompt": "Write a Harbour Base64 encoder/decode function. Use Asc(), Chr(), and bit operations.",
"expected_keywords": ["FUNCTION", "LOCAL", "Asc", "Chr"],
"min_lines": 10,
},
{
"id": "CMPX_06", "category": "Complex", "name": "JSON serializer",
"prompt": "Write a Harbour function that serializes a hash to JSON string. Handle strings, numbers, booleans, arrays, and nested objects.",
"expected_keywords": ["FUNCTION", "LOCAL", "HB_IsHash"],
"min_lines": 15,
},
{
"id": "CMPX_07", "category": "Complex", "name": "LRU Cache",
"prompt": "Write a Harbour LRU cache class with get/set/delete, TTL expiration, and max size. Use a hash and an array for ordering.",
"expected_keywords": ["CLASS", "DATA", "METHOD"],
"min_lines": 15,
},
{
"id": "CMPX_08", "category": "Complex", "name": "SQL-like query on arrays",
"prompt": "Write a Harbour function that filters an array of hashes like SQL WHERE clause. Support =, <>, >, <, LIKE operators.",
"expected_keywords": ["FUNCTION", "LOCAL", "FOR"],
"min_lines": 12,
},
{
"id": "CMPX_09", "category": "Complex", "name": "Rate limiter",
"prompt": "Write a Harbour rate limiter class: max N requests per M seconds. Use timestamps and a queue.",
"expected_keywords": ["CLASS", "METHOD", "LOCAL"],
"min_lines": 12,
},
{
"id": "CMPX_10", "category": "Complex", "name": "Config file writer",
"prompt": "Write a Harbour config manager that saves/loads settings to JSON file. Include defaults, validation, and typed getters.",
"expected_keywords": ["FUNCTION", "LOCAL", "FClose"],
"min_lines": 12,
},
# ---- BUGGY CODE TO FIX ----
{
"id": "FIX_01", "category": "Bug Fix", "name": "Null pointer",
"prompt": "Fix this Harbour code that crashes when array is empty:\nLOCAL a := {}\n? a[1]",
"expected_keywords": ["IF", "LEN", "RETURN"],
"min_lines": 3,
},
{
"id": "FIX_02", "category": "Bug Fix", "name": "Wrong loop bounds",
"prompt": "Fix this code that skips last element:\nLOCAL a := {10,20,30}\nFOR i := 1 TO LEN(a)-1\n ? a[i]\nNEXT",
"expected_keywords": ["FOR", "TO", "LEN"],
"min_lines": 3,
},
{
"id": "FIX_03", "category": "Bug Fix", "name": "String concat error",
"prompt": "Fix this code that fails on nil values:\nLOCAL cName := NIL\n? 'Hello ' + cName",
"expected_keywords": ["IF", "LOCAL", "RETURN"],
"min_lines": 3,
},
# ---- HARBOUR-SPECIFIC ----
{
"id": "HARB_01", "category": "Harbour-Specific", "name": "HB_* functions",
"prompt": "Write Harbour code using HB_IsString, HB_IsNumeric, HB_IsArray, HB_IsHash, HB_IsNil to validate function arguments. Include proper error messages.",
"expected_keywords": ["HB_IsString", "HB_IsNumeric", "IF"],
"min_lines": 6,
},
{
"id": "HARB_02", "category": "Harbour-Specific", "name": "Regex",
"prompt": "Write a Harbour function using HB_RegEx to validate email addresses. Use HB_RegExCompile and HB_RegExMatch.",
"expected_keywords": ["HB_RegEx", "FUNCTION"],
"min_lines": 6,
},
{
"id": "HARB_03", "category": "Harbour-Specific", "name": "Serialization",
"prompt": "Write Harbour code that serializes a hash to binary with HB_Serialize and deserializes with HB_Deserialize.",
"expected_keywords": ["HB_Serialize", "HB_Deserialize"],
"min_lines": 6,
},
{
"id": "HARB_04", "category": "Harbour-Specific", "name": "File path operations",
"prompt": "Write Harbour code using hb_DirBuild, hb_DirNameGet, hb_FileNameGet, hb_PathNormalize for cross-platform file handling.",
"expected_keywords": ["hb_Dir", "hb_File", "hb_Path"],
"min_lines": 6,
},
]
# ============================================================
# MAIN
# ============================================================
def main():
print("=" * 70)
print("HARBOUR CODE GENERATION TEST BATTERY")
print(f"Model: {MODEL}")
print(f"Tests: {len(TESTS)}")
print(f"Harbour: {HARBOUR}")
print(f"Started: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
print("=" * 70)
SYSTEM = """You are an expert Harbour programmer. Write clean, correct, COMPILABLE Harbour code.
Use Hungarian notation: n=numeric, c=character, l=logical, a=array, o=object, d=date.
Use 3-space indentation.
Do NOT include explanations or markdown. Only raw Harbour code.
End functions with RETURN and END FUNCTION."""
results = []
compile_pass = 0
compile_fail = 0
for i, test in enumerate(TESTS, 1):
print(f"\n[{i:2d}/{len(TESTS)}] {test['id']}: {test['name']}")
# Query model
result = query_ollama(test["prompt"], SYSTEM)
if result["error"]:
print(f" MODEL ERROR: {result['error']}")
results.append({"test": test, "model_error": result["error"], "compile": False, "compile_error": ""})
continue
# Clean response
code = clean_code(result["response"])
# Check for expected keywords
keywords_found = [kw for kw in test["expected_keywords"] if kw.upper() in code.upper()]
keywords_missing = [kw for kw in test["expected_keywords"] if kw.upper() not in code.upper()]
# Compile
success, error, obj = compile_harbour(code)
status = "PASS" if success else "FAIL"
if success:
compile_pass += 1
else:
compile_fail += 1
print(f" Compile: {status} | Keywords: {len(keywords_found)}/{len(test['expected_keywords'])} | TPS: {result['tps']:.0f}")
if keywords_missing:
print(f" Missing keywords: {', '.join(keywords_missing)}")
if not success and error:
# Show first error only
first_error = error.split('\n')[0][:120]
print(f" Error: {first_error}")
results.append({
"test": test,
"code": code[:3000],
"compile_success": success,
"compile_error": error[:500] if error else "",
"keywords_found": keywords_found,
"keywords_missing": keywords_missing,
"tokens": result["eval_count"],
"tps": result["tps"],
"duration": result["duration"],
"lines": code.count('\n') + 1,
})
# Summary by category
print("\n" + "=" * 70)
print("RESULTS SUMMARY")
print("=" * 70)
categories = {}
for r in results:
cat = r["test"]["category"]
if cat not in categories:
categories[cat] = {"pass": 0, "fail": 0, "total": 0}
categories[cat]["total"] += 1
if r.get("compile_success"):
categories[cat]["pass"] += 1
else:
categories[cat]["fail"] += 1
print(f"\n{'Category':<20} {'Pass':<6} {'Fail':<6} {'Rate':<8}")
print("-" * 45)
for cat, data in sorted(categories.items()):
rate = data["pass"] / data["total"] * 100 if data["total"] > 0 else 0
print(f"{cat:<20} {data['pass']:<6} {data['fail']:<6} {rate:.0f}%")
print(f"\n{'TOTAL':<20} {compile_pass:<6} {compile_fail:<6} {compile_pass/len(results)*100:.0f}%")
print(f"Total tests: {len(results)}")
total_tokens = sum(r.get("tokens", 0) for r in results)
total_time = sum(r.get("duration", 0) for r in results)
print(f"Total tokens: {total_tokens:,}")
print(f"Total time: {total_time:.1f}s")
# Save
output = Path("/home/fivetech/finetune/test_baseline_qwen36.json")
with open(output, "w") as f:
json.dump({
"model": MODEL,
"timestamp": datetime.now().isoformat(),
"compile_pass": compile_pass,
"compile_fail": compile_fail,
"compile_rate": compile_pass / len(results) * 100,
"categories": categories,
"results": results,
}, f, indent=2, ensure_ascii=False)
print(f"\nResults saved to: {output}")
if __name__ == "__main__":
main()
|