Text Generation
Transformers
Safetensors
GGUF
Korean
English
llama
3b
korean
from-scratch
orpo
instruction-tuned
preference-aligned
fp8
b200
Eval Results (legacy)
text-generation-inference
Instructions to use pathcosmos/frankenstallm with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use pathcosmos/frankenstallm with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="pathcosmos/frankenstallm")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("pathcosmos/frankenstallm") model = AutoModelForCausalLM.from_pretrained("pathcosmos/frankenstallm", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use pathcosmos/frankenstallm 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 pathcosmos/frankenstallm:Q4_K_M # Run inference directly in the terminal: llama cli -hf pathcosmos/frankenstallm:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf pathcosmos/frankenstallm:Q4_K_M # Run inference directly in the terminal: llama cli -hf pathcosmos/frankenstallm: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 pathcosmos/frankenstallm:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf pathcosmos/frankenstallm: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 pathcosmos/frankenstallm:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf pathcosmos/frankenstallm:Q4_K_M
Use Docker
docker model run hf.co/pathcosmos/frankenstallm:Q4_K_M
- LM Studio
- Jan
- vLLM
How to use pathcosmos/frankenstallm with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "pathcosmos/frankenstallm" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "pathcosmos/frankenstallm", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/pathcosmos/frankenstallm:Q4_K_M
- SGLang
How to use pathcosmos/frankenstallm 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 "pathcosmos/frankenstallm" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "pathcosmos/frankenstallm", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'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 "pathcosmos/frankenstallm" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "pathcosmos/frankenstallm", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Ollama
How to use pathcosmos/frankenstallm with Ollama:
ollama run hf.co/pathcosmos/frankenstallm:Q4_K_M
- Unsloth Desktop
- Docker Model Runner
How to use pathcosmos/frankenstallm with Docker Model Runner:
docker model run hf.co/pathcosmos/frankenstallm:Q4_K_M
- Lemonade
How to use pathcosmos/frankenstallm with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull pathcosmos/frankenstallm:Q4_K_M
Run and chat with the model
lemonade run user.frankenstallm-Q4_K_M
List all available models
lemonade list
- Atomic Chat
| # ============================================================================= | |
| # ORPO Training Completion Watchdog | |
| # ============================================================================= | |
| # Monitors the ORPO training process. When it finishes, automatically launches | |
| # the comprehensive evaluation pipeline. | |
| # | |
| # Usage: | |
| # nohup bash scripts/orpo_eval_watchdog.sh > checkpoints/korean_3b_orpo_v1/watchdog.log 2>&1 & | |
| # ============================================================================= | |
| set -euo pipefail | |
| PROJECT_ROOT="/PROJECT/0325120031_A/ghong/taketimes/llm-bang" | |
| TRAIN_LOG="${PROJECT_ROOT}/checkpoints/korean_3b_orpo_v1/train.log" | |
| TRAIN_PID=$(pgrep -f "train/orpo.py.*korean_3b_orpo_v1" | head -1) | |
| echo "==============================================" | |
| echo " ORPO Eval Watchdog Started" | |
| echo "==============================================" | |
| echo " Time : $(date '+%Y-%m-%d %H:%M:%S')" | |
| echo " Train PID : ${TRAIN_PID:-NOT FOUND}" | |
| echo " Train Log : ${TRAIN_LOG}" | |
| echo "==============================================" | |
| if [ -z "${TRAIN_PID}" ]; then | |
| echo "[WARN] Training process not found. Checking if already completed..." | |
| # Check if training already finished by looking for final output | |
| if grep -q "Training completed" "${TRAIN_LOG}" 2>/dev/null || \ | |
| grep -q "Saving model checkpoint" "${TRAIN_LOG}" 2>/dev/null; then | |
| echo "[INFO] Training appears to have already completed." | |
| else | |
| echo "[ERROR] No training process and no completion marker found. Exiting." | |
| exit 1 | |
| fi | |
| else | |
| echo "[INFO] Watching training PID ${TRAIN_PID}..." | |
| echo "" | |
| # Poll every 60 seconds | |
| while kill -0 "${TRAIN_PID}" 2>/dev/null; do | |
| # Get current step | |
| CURRENT_STEP=$(grep -oP '\d+/9840' "${TRAIN_LOG}" 2>/dev/null | tail -1 || echo "?/?") | |
| LATEST_LOSS=$(grep "'loss':" "${TRAIN_LOG}" 2>/dev/null | tail -1 | grep -oP "'loss': '([^']+)'" | sed "s/'loss': '//;s/'//" || echo "?") | |
| echo "[$(date '+%H:%M:%S')] Step ${CURRENT_STEP} | Loss: ${LATEST_LOSS} | PID ${TRAIN_PID} running" | |
| sleep 60 | |
| done | |
| echo "" | |
| echo "==============================================" | |
| echo "[INFO] Training process ${TRAIN_PID} has ended." | |
| echo "[INFO] Time: $(date '+%Y-%m-%d %H:%M:%S')" | |
| echo "==============================================" | |
| fi | |
| # Wait a moment for any final I/O | |
| sleep 10 | |
| # Get final training stats | |
| echo "" | |
| echo "[INFO] Final training stats:" | |
| grep "eval_loss" "${TRAIN_LOG}" | tail -1 | tr ',' '\n' | head -10 | |
| echo "" | |
| # Detect the latest checkpoint | |
| LATEST_CKPT=$(ls -d ${PROJECT_ROOT}/checkpoints/korean_3b_orpo_v1/checkpoint-* 2>/dev/null | sort -t- -k2 -n | tail -1) | |
| echo "[INFO] Latest checkpoint: ${LATEST_CKPT}" | |
| if [ -z "${LATEST_CKPT}" ]; then | |
| echo "[ERROR] No checkpoint found. Cannot proceed with evaluation." | |
| exit 1 | |
| fi | |
| # Send telegram notification (if available) | |
| python3 -c " | |
| import os, urllib.request, urllib.parse, json | |
| token = os.environ.get('TELEGRAM_BOT_TOKEN', '') | |
| chat_id = os.environ.get('TELEGRAM_CHAT_ID', '') | |
| if token and chat_id: | |
| msg = 'π ORPO νμ΅ μλ£! μλ νκ° μμν©λλ€.\nCheckpoint: ${LATEST_CKPT##*/}' | |
| url = f'https://api.telegram.org/bot{token}/sendMessage' | |
| data = urllib.parse.urlencode({'chat_id': chat_id, 'text': msg}).encode() | |
| urllib.request.urlopen(url, data, timeout=10) | |
| print('[INFO] Telegram notification sent.') | |
| else: | |
| print('[INFO] Telegram not configured, skipping notification.') | |
| " 2>/dev/null || true | |
| # ============================================================================ | |
| # Launch evaluation pipeline | |
| # ============================================================================ | |
| echo "" | |
| echo "==============================================" | |
| echo " Starting ORPO Evaluation Pipeline" | |
| echo " Time: $(date '+%Y-%m-%d %H:%M:%S')" | |
| echo "==============================================" | |
| cd "${PROJECT_ROOT}" | |
| python3 eval/orpo_eval_pipeline.py \ | |
| --checkpoint "${LATEST_CKPT}" \ | |
| 2>&1 | tee -a checkpoints/korean_3b_orpo_v1/eval.log | |
| EVAL_EXIT=$? | |
| echo "" | |
| echo "==============================================" | |
| echo " Evaluation Complete" | |
| echo " Exit code: ${EVAL_EXIT}" | |
| echo " Time: $(date '+%Y-%m-%d %H:%M:%S')" | |
| echo "==============================================" | |
| # Send completion notification | |
| python3 -c " | |
| import os, urllib.request, urllib.parse | |
| token = os.environ.get('TELEGRAM_BOT_TOKEN', '') | |
| chat_id = os.environ.get('TELEGRAM_CHAT_ID', '') | |
| if token and chat_id: | |
| exit_code = ${EVAL_EXIT} | |
| status = 'β μ±κ³΅' if exit_code == 0 else 'β μ€ν¨' | |
| msg = f'ORPO νκ° μλ£: {status}\nExit code: {exit_code}\nλ³΄κ³ μ: reports/ νμΈ' | |
| url = f'https://api.telegram.org/bot{token}/sendMessage' | |
| data = urllib.parse.urlencode({'chat_id': chat_id, 'text': msg}).encode() | |
| urllib.request.urlopen(url, data, timeout=10) | |
| " 2>/dev/null || true | |
| exit ${EVAL_EXIT} | |