Text Generation
Transformers
English
qwen2
code-generation
python
fine-tuning
Qwen
tools
agent-framework
multi-agent
conversational
Eval Results (legacy)
Instructions to use my-ai-stack/Stack-2-9-finetuned with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use my-ai-stack/Stack-2-9-finetuned with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="my-ai-stack/Stack-2-9-finetuned") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("my-ai-stack/Stack-2-9-finetuned") model = AutoModelForCausalLM.from_pretrained("my-ai-stack/Stack-2-9-finetuned", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use my-ai-stack/Stack-2-9-finetuned with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "my-ai-stack/Stack-2-9-finetuned" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "my-ai-stack/Stack-2-9-finetuned", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/my-ai-stack/Stack-2-9-finetuned
- SGLang
How to use my-ai-stack/Stack-2-9-finetuned 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 "my-ai-stack/Stack-2-9-finetuned" \ --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": "my-ai-stack/Stack-2-9-finetuned", "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 "my-ai-stack/Stack-2-9-finetuned" \ --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": "my-ai-stack/Stack-2-9-finetuned", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use my-ai-stack/Stack-2-9-finetuned with Docker Model Runner:
docker model run hf.co/my-ai-stack/Stack-2-9-finetuned
| # ============================================================================= | |
| # Stack 2.9 Full Benchmark Evaluation Suite | |
| # ============================================================================= | |
| # Runs all benchmarks and generates comprehensive evaluation report. | |
| # | |
| # Usage: | |
| # ./run_all_benchmarks.sh [OPTIONS] | |
| # | |
| # Options: | |
| # --model MODEL Model name to evaluate (default: stack-2.9) | |
| # --output DIR Output directory (default: ./results) | |
| # --skip-slow Skip slow benchmarks | |
| # --sample-size N Use N samples per benchmark (default: all) | |
| # --verbose Verbose output | |
| # | |
| # ============================================================================= | |
| set -e | |
| # Configuration | |
| MODEL="${MODEL:-stack-2.9}" | |
| OUTPUT_DIR="${OUTPUT_DIR:-./results}" | |
| SAMPLE_SIZE="" | |
| SKIP_SLOW="" | |
| VERBOSE="" | |
| PYTHON="${PYTHON:-python3}" | |
| # Colors | |
| RED='\033[0;31m' | |
| GREEN='\033[0;32m' | |
| YELLOW='\033[1;33m' | |
| BLUE='\033[0;34m' | |
| NC='\033[0m' # No Color | |
| # Benchmark scripts | |
| SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" | |
| HUMAN_EVAL="${SCRIPT_DIR}/human_eval.py" | |
| MBPP_EVAL="${SCRIPT_DIR}/mbpp_eval.py" | |
| TOOL_EVAL="${SCRIPT_DIR}/tool_use_eval.py" | |
| SELF_IMPROVE_EVAL="${SCRIPT_DIR}/self_improve_eval.py" | |
| DASHBOARD="${SCRIPT_DIR}/results/dashboard.py" | |
| # ============================================================================= | |
| # Helper Functions | |
| # ============================================================================= | |
| log_info() { | |
| echo -e "${BLUE}[INFO]${NC} $1" | |
| } | |
| log_success() { | |
| echo -e "${GREEN}[SUCCESS]${NC} $1" | |
| } | |
| log_warning() { | |
| echo -e "${YELLOW}[WARNING]${NC} $1" | |
| } | |
| log_error() { | |
| echo -e "${RED}[ERROR]${NC} $1" | |
| } | |
| section() { | |
| echo "" | |
| echo "==============================================================================" | |
| echo "$1" | |
| echo "==============================================================================" | |
| } | |
| # ============================================================================= | |
| # Parse Arguments | |
| # ============================================================================= | |
| while [[ $# -gt 0 ]]; do | |
| case $1 in | |
| --model) | |
| MODEL="$2" | |
| shift 2 | |
| ;; | |
| --output) | |
| OUTPUT_DIR="$2" | |
| shift 2 | |
| ;; | |
| --skip-slow) | |
| SKIP_SLOW="1" | |
| shift | |
| ;; | |
| --sample-size) | |
| SAMPLE_SIZE="$2" | |
| shift 2 | |
| ;; | |
| --verbose) | |
| VERBOSE="1" | |
| shift | |
| ;; | |
| --help) | |
| echo "Stack 2.9 Full Benchmark Suite" | |
| echo "" | |
| echo "Usage: $0 [OPTIONS]" | |
| echo "" | |
| echo "Options:" | |
| echo " --model MODEL Model name (default: stack-2.9)" | |
| echo " --output DIR Output directory (default: ./results)" | |
| echo " --skip-slow Skip slow benchmarks" | |
| echo " --sample-size N Sample size for each benchmark" | |
| echo " --verbose Verbose output" | |
| echo " --help Show this help message" | |
| exit 0 | |
| ;; | |
| *) | |
| log_error "Unknown option: $1" | |
| exit 1 | |
| ;; | |
| esac | |
| done | |
| # ============================================================================= | |
| # Setup | |
| # ============================================================================= | |
| log_info "Stack 2.9 Benchmark Suite" | |
| log_info "Model: ${MODEL}" | |
| log_info "Output: ${OUTPUT_DIR}" | |
| echo "" | |
| # Create output directory | |
| mkdir -p "${OUTPUT_DIR}" | |
| mkdir -p "${OUTPUT_DIR}/detailed" | |
| # Track start time | |
| START_TIME=$(date +%s) | |
| # Results summary | |
| declare -A BENCHMARK_RESULTS | |
| # ============================================================================= | |
| # Check Dependencies | |
| # ============================================================================= | |
| section "Checking Dependencies" | |
| check_python() { | |
| if command -v python3 &> /dev/null; then | |
| PYTHON="python3" | |
| elif command -v python &> /dev/null; then | |
| PYTHON="python" | |
| else | |
| log_error "Python not found!" | |
| exit 1 | |
| fi | |
| log_success "Python: $(${PYTHON} --version)" | |
| } | |
| check_dependencies() { | |
| log_info "Checking Python dependencies..." | |
| # Check for required modules | |
| REQUIRED_MODULES=("json" "datetime" "pathlib" "argparse") | |
| MISSING="" | |
| for module in "${REQUIRED_MODULES[@]}"; do | |
| if ! ${PYTHON} -c "import ${module}" &> /dev/null; then | |
| MISSING="${MISSING} ${module}" | |
| fi | |
| done | |
| if [ -n "${MISSING}" ]; then | |
| log_warning "Missing modules:${MISSING}" | |
| log_info "These are standard library modules and should be available." | |
| fi | |
| log_success "Dependencies OK" | |
| } | |
| check_python | |
| check_dependencies | |
| # ============================================================================= | |
| # HumanEval Benchmark | |
| # ============================================================================= | |
| section "HumanEval Benchmark" | |
| log_info "Running HumanEval benchmark..." | |
| log_info "Metrics: Pass@1, Pass@10, Pass@100" | |
| HUMAN_EVAL_START=$(date +%s) | |
| if [ -f "${HUMAN_EVAL}" ]; then | |
| HUMAN_EVAL_CMD="${PYTHON} ${HUMAN_EVAL} --model ${MODEL} --output ${OUTPUT_DIR}/detailed" | |
| if [ -n "${SAMPLE_SIZE}" ]; then | |
| # Note: human_eval.py doesn't support sample-size directly | |
| # but we include it for other benchmarks | |
| : | |
| fi | |
| if [ -n "${VERBOSE}" ]; then | |
| ${HUMAN_EVAL_CMD} 2>&1 | tee "${OUTPUT_DIR}/detailed/humaneval_output.log" | |
| else | |
| ${HUMAN_EVAL_CMD} > "${OUTPUT_DIR}/detailed/humaneval_output.log" 2>&1 | |
| fi | |
| HUMAN_EVAL_END=$(date +%s) | |
| HUMAN_EVAL_TIME=$((HUMAN_EVAL_END - HUMAN_EVAL_START)) | |
| if [ -f "${OUTPUT_DIR}/detailed/humaneval_results.json" ]; then | |
| PASS_1=$(grep -o '"pass_at_1": [0-9.]*' "${OUTPUT_DIR}/detailed/humaneval_results.json" | cut -d':' -f2) | |
| PASS_10=$(grep -o '"pass_at_10": [0-9.]*' "${OUTPUT_DIR}/detailed/humaneval_results.json" | cut -d':' -f2) | |
| BENCHMARK_RESULTS["humaneval_pass1"]="${PASS_1}" | |
| BENCHMARK_RESULTS["humaneval_pass10"]="${PASS_10}" | |
| log_success "HumanEval: Pass@1=${PASS_1}, Pass@10=${PASS_10} (${HUMAN_EVAL_TIME}s)" | |
| else | |
| log_error "HumanEval results not found" | |
| fi | |
| else | |
| log_warning "HumanEval script not found: ${HUMAN_EVAL}" | |
| fi | |
| # ============================================================================= | |
| # MBPP Benchmark | |
| # ============================================================================= | |
| section "MBPP Benchmark" | |
| log_info "Running MBPP benchmark..." | |
| log_info "Metrics: Pass@1, Pass@10" | |
| MBPP_START=$(date +%s) | |
| if [ -f "${MBPP_EVAL}" ]; then | |
| MBPP_CMD="${PYTHON} ${MBPP_EVAL} --model ${MODEL} --output ${OUTPUT_DIR}/detailed" | |
| if [ -n "${VERBOSE}" ]; then | |
| ${MBPP_CMD} 2>&1 | tee "${OUTPUT_DIR}/detailed/mbpp_output.log" | |
| else | |
| ${MBPP_CMD} > "${OUTPUT_DIR}/detailed/mbpp_output.log" 2>&1 | |
| fi | |
| MBPP_END=$(date +%s) | |
| MBPP_TIME=$((MBPP_END - MBPP_START)) | |
| if [ -f "${OUTPUT_DIR}/detailed/mbpp_results.json" ]; then | |
| PASS_1=$(grep -o '"pass_at_1": [0-9.]*' "${OUTPUT_DIR}/detailed/mbpp_results.json" | cut -d':' -f2) | |
| PASS_10=$(grep -o '"pass_at_10": [0-9.]*' "${OUTPUT_DIR}/detailed/mbpp_results.json" | cut -d':' -f2) | |
| BENCHMARK_RESULTS["mbpp_pass1"]="${PASS_1}" | |
| BENCHMARK_RESULTS["mbpp_pass10"]="${PASS_10}" | |
| log_success "MBPP: Pass@1=${PASS_1}, Pass@10=${PASS_10} (${MBPP_TIME}s)" | |
| else | |
| log_error "MBPP results not found" | |
| fi | |
| else | |
| log_warning "MBPP script not found: ${MBPP_EVAL}" | |
| fi | |
| # ============================================================================= | |
| # Tool Use Evaluation | |
| # ============================================================================= | |
| section "Tool Use Evaluation" | |
| log_info "Running Tool Use evaluation..." | |
| log_info "Metrics: Tool Selection Accuracy, Parameter Accuracy, Execution Success" | |
| TOOL_START=$(date +%s) | |
| if [ -f "${TOOL_EVAL}" ]; then | |
| TOOL_CMD="${PYTHON} ${TOOL_EVAL} --model ${MODEL} --output ${OUTPUT_DIR}/detailed" | |
| if [ -n "${SAMPLE_SIZE}" ]; then | |
| TOOL_CMD="${TOOL_CMD} --sample ${SAMPLE_SIZE}" | |
| fi | |
| if [ -n "${VERBOSE}" ]; then | |
| ${TOOL_CMD} 2>&1 | tee "${OUTPUT_DIR}/detailed/tool_output.log" | |
| else | |
| ${TOOL_CMD} > "${OUTPUT_DIR}/detailed/tool_output.log" 2>&1 | |
| fi | |
| TOOL_END=$(date +%s) | |
| TOOL_TIME=$((TOOL_END - TOOL_START)) | |
| if [ -f "${OUTPUT_DIR}/detailed/tool_use_results.json" ]; then | |
| TOOL_ACC=$(grep -o '"tool_selection_accuracy": [0-9.]*' "${OUTPUT_DIR}/detailed/tool_use_results.json" | cut -d':' -f2) | |
| PARAM_ACC=$(grep -o '"parameter_accuracy": [0-9.]*' "${OUTPUT_DIR}/detailed/tool_use_results.json" | cut -d':' -f2) | |
| EXEC_RATE=$(grep -o '"execution_success_rate": [0-9.]*' "${OUTPUT_DIR}/detailed/tool_use_results.json" | cut -d':' -f2) | |
| BENCHMARK_RESULTS["tool_selection_accuracy"]="${TOOL_ACC}" | |
| BENCHMARK_RESULTS["parameter_accuracy"]="${PARAM_ACC}" | |
| BENCHMARK_RESULTS["execution_success_rate"]="${EXEC_RATE}" | |
| log_success "Tool Use: Selection=${TOOL_ACC}, Param=${PARAM_ACC}, Exec=${EXEC_RATE} (${TOOL_TIME}s)" | |
| else | |
| log_error "Tool Use results not found" | |
| fi | |
| else | |
| log_warning "Tool Use script not found: ${TOOL_EVAL}" | |
| fi | |
| # ============================================================================= | |
| # Self-Improvement Evaluation | |
| # ============================================================================= | |
| if [ -z "${SKIP_SLOW}" ]; then | |
| section "Self-Improvement Evaluation" | |
| log_info "Running Self-Improvement evaluation..." | |
| log_info "Metrics: Memory Retention, Pattern Application, Improvement Rate" | |
| SELF_IMPROVE_START=$(date +%s) | |
| if [ -f "${SELF_IMPROVE_EVAL}" ]; then | |
| SELF_CMD="${PYTHON} ${SELF_IMPROVE_EVAL} --model ${MODEL} --output ${OUTPUT_DIR}/detailed" | |
| if [ -n "${VERBOSE}" ]; then | |
| ${SELF_CMD} 2>&1 | tee "${OUTPUT_DIR}/detailed/self_improve_output.log" | |
| else | |
| ${SELF_CMD} > "${OUTPUT_DIR}/detailed/self_improve_output.log" 2>&1 | |
| fi | |
| SELF_IMPROVE_END=$(date +%s) | |
| SELF_TIME=$((SELF_IMPROVE_END - SELF_IMPROVE_START)) | |
| if [ -f "${OUTPUT_DIR}/detailed/self_improve_results.json" ]; then | |
| MEM_RET=$(grep -o '"memory_retention_rate": [0-9.]*' "${OUTPUT_DIR}/detailed/self_improve_results.json" | cut -d':' -f2) | |
| PATTERN_ACC=$(grep -o '"pattern_application_accuracy": [0-9.]*' "${OUTPUT_DIR}/detailed/self_improve_results.json" | cut -d':' -f2) | |
| IMPROVE_RATE=$(grep -o '"improvement_rate": [0-9.]*' "${OUTPUT_DIR}/detailed/self_improve_results.json" | cut -d':' -f2) | |
| BENCHMARK_RESULTS["memory_retention"]="${MEM_RET}" | |
| BENCHMARK_RESULTS["pattern_accuracy"]="${PATTERN_ACC}" | |
| BENCHMARK_RESULTS["improvement_rate"]="${IMPROVE_RATE}" | |
| log_success "Self-Improve: Memory=${MEM_RET}, Pattern=${PATTERN_ACC}, Improve=${IMPROVE_RATE} (${SELF_TIME}s)" | |
| else | |
| log_error "Self-Improvement results not found" | |
| fi | |
| else | |
| log_warning "Self-Improvement script not found: ${SELF_IMPROVE_EVAL}" | |
| fi | |
| else | |
| log_info "Skipping Self-Improvement evaluation (--skip-slow)" | |
| fi | |
| # ============================================================================= | |
| # Generate Dashboard | |
| # ============================================================================= | |
| section "Generating Dashboard" | |
| log_info "Creating visualization dashboard..." | |
| if [ -f "${DASHBOARD}" ]; then | |
| ${PYTHON} "${DASHBOARD}" --results-dir "${OUTPUT_DIR}/detailed" --output "${OUTPUT_DIR}" 2>&1 | tee "${OUTPUT_DIR}/detailed/dashboard_output.log" | |
| log_success "Dashboard generated at ${OUTPUT_DIR}/dashboard.html" | |
| else | |
| log_warning "Dashboard script not found: ${DASHBOARD}" | |
| fi | |
| # ============================================================================= | |
| # Generate Summary Report | |
| # ============================================================================= | |
| section "Summary Report" | |
| TOTAL_TIME=$(($(date +%s) - START_TIME)) | |
| echo "" | |
| echo "==============================================================================" | |
| echo "BENCHMARK RESULTS SUMMARY" | |
| echo "==============================================================================" | |
| echo "" | |
| echo "Model: ${MODEL}" | |
| echo "Evaluation Date: $(date '+%Y-%m-%d %H:%M:%S')" | |
| echo "Total Time: ${TOTAL_TIME}s" | |
| echo "" | |
| echo "------------------------------------------------------------------------------" | |
| echo "CODE GENERATION BENCHMARKS" | |
| echo "------------------------------------------------------------------------------" | |
| printf "%-20s %-15s %-15s\n" "Benchmark" "Pass@1" "Pass@10" | |
| echo "------------------------------------------------------------------------------" | |
| printf "%-20s %-15s %-15s\n" "HumanEval" "${BENCHMARK_RESULTS[humaneval_pass1]:-N/A}" "${BENCHMARK_RESULTS[humaneval_pass10]:-N/A}" | |
| printf "%-20s %-15s %-15s\n" "MBPP" "${BENCHMARK_RESULTS[mbpp_pass1]:-N/A}" "${BENCHMARK_RESULTS[mbpp_pass10]:-N/A}" | |
| echo "" | |
| echo "------------------------------------------------------------------------------" | |
| echo "TOOL USE CAPABILITIES" | |
| echo "------------------------------------------------------------------------------" | |
| printf "%-25s %-15s\n" "Metric" "Value" | |
| echo "------------------------------------------------------------------------------" | |
| printf "%-25s %-15s\n" "Tool Selection Accuracy" "${BENCHMARK_RESULTS[tool_selection_accuracy]:-N/A}" | |
| printf "%-25s %-15s\n" "Parameter Accuracy" "${BENCHMARK_RESULTS[parameter_accuracy]:-N/A}" | |
| printf "%-25s %-15s\n" "Execution Success Rate" "${BENCHMARK_RESULTS[execution_success_rate]:-N/A}" | |
| echo "" | |
| echo "------------------------------------------------------------------------------" | |
| echo "SELF-IMPROVEMENT CAPABILITIES" | |
| echo "------------------------------------------------------------------------------" | |
| printf "%-25s %-15s\n" "Metric" "Value" | |
| echo "------------------------------------------------------------------------------" | |
| printf "%-25s %-15s\n" "Memory Retention Rate" "${BENCHMARK_RESULTS[memory_retention]:-N/A}" | |
| printf "%-25s %-15s\n" "Pattern Application Accuracy" "${BENCHMARK_RESULTS[pattern_accuracy]:-N/A}" | |
| printf "%-25s %-15s\n" "Improvement Rate" "${BENCHMARK_RESULTS[improvement_rate]:-N/A}" | |
| echo "" | |
| echo "==============================================================================" | |
| # ============================================================================= | |
| # Save Summary to JSON | |
| # ============================================================================= | |
| cat > "${OUTPUT_DIR}/benchmark_summary.json" << EOF | |
| { | |
| "model": "${MODEL}", | |
| "evaluation_date": "$(date '+%Y-%m-%d %H:%M:%S')", | |
| "total_time_seconds": ${TOTAL_TIME}, | |
| "humaneval": { | |
| "pass_at_1": ${BENCHMARK_RESULTS[humaneval_pass1]:-null}, | |
| "pass_at_10": ${BENCHMARK_RESULTS[humaneval_pass10]:-null} | |
| }, | |
| "mbpp": { | |
| "pass_at_1": ${BENCHMARK_RESULTS[mbpp_pass1]:-null}, | |
| "pass_at_10": ${BENCHMARK_RESULTS[mbpp_pass10]:-null} | |
| }, | |
| "tool_use": { | |
| "tool_selection_accuracy": ${BENCHMARK_RESULTS[tool_selection_accuracy]:-null}, | |
| "parameter_accuracy": ${BENCHMARK_RESULTS[parameter_accuracy]:-null}, | |
| "execution_success_rate": ${BENCHMARK_RESULTS[execution_success_rate]:-null} | |
| }, | |
| "self_improvement": { | |
| "memory_retention_rate": ${BENCHMARK_RESULTS[memory_retention]:-null}, | |
| "pattern_application_accuracy": ${BENCHMARK_RESULTS[pattern_accuracy]:-null}, | |
| "improvement_rate": ${BENCHMARK_RESULTS[improvement_rate]:-null} | |
| } | |
| } | |
| EOF | |
| log_success "Summary saved to ${OUTPUT_DIR}/benchmark_summary.json" | |
| log_success "Detailed results in ${OUTPUT_DIR}/detailed/" | |
| echo "" | |
| log_success "All benchmarks completed successfully!" | |