laptopLLM/Makefile at master · d3bvstack-archive/laptopLLM · GitHub

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
.PHONY: up down rebuild logs shell pull-model clean status health test-chat gpu-test

COMPOSE_FILE := docker-compose.yml
ENV_FILE := .env

up: pull-model
	docker compose --env-file $(ENV_FILE) -f $(COMPOSE_FILE) up -d

down:
	docker compose --env-file $(ENV_FILE) -f $(COMPOSE_FILE) down

rebuild: down up

logs:
	docker compose --env-file $(ENV_FILE) -f $(COMPOSE_FILE) logs -f --tail=100

shell:
	docker compose --env-file $(ENV_FILE) -f $(COMPOSE_FILE) exec llama-server sh

pull-model:
	@mkdir -p models
	@MODEL_FILE=$$(grep MODEL_FILE $(ENV_FILE) | cut -d'=' -f2); \
	if [ -f "models/$$MODEL_FILE" ]; then \
		echo "Model already exists: models/$$MODEL_FILE"; \
	else \
		echo "Downloading model: $$MODEL_FILE"; \
		if command -v huggingface-cli >/dev/null 2>&1; then \
			huggingface-cli download LiquidAI/LFM2.5-1.2B-Instruct-GGUF \
				--include "$$MODEL_FILE" \
				--local-dir models --local-dir-use-symlinks False; \
		elif command -v wget >/dev/null 2>&1; then \
			wget -P models "https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct-GGUF/resolve/main/$$MODEL_FILE"; \
		else \
			echo "Error: Neither huggingface-cli nor wget found. Install one to download models."; \
			exit 1; \
		fi; \
	fi

clean:
	docker compose --env-file $(ENV_FILE) -f $(COMPOSE_FILE) down -v --remove-orphans
	docker image prune -f
	rm -rf models/*

status:
	docker compose --env-file $(ENV_FILE) -f $(COMPOSE_FILE) ps

health:
	@echo "Checking server health..."
	@curl -sf http://localhost:$(shell grep HOST_PORT $(ENV_FILE) | cut -d'=' -f2)/health && echo "Server healthy" || echo "Server unhealthy"

test-chat:
	@curl -s http://localhost:$(shell grep HOST_PORT $(ENV_FILE) | cut -d'=' -f2)/v1/chat/completions \
		-H "Content-Type: application/json" \
		-d '{"model": "lfm2.5-1.2b-instruct", "messages": [{"role": "user", "content": "Hello!"}], "max_tokens": 50}' | jq .

gpu-test:
	@echo "Testing GPU detection in container..."
	@docker run --rm --gpus all nvidia/cuda:12.6-base nvidia-smi 2>&1 || echo "nvidia-smi failed - check nvidia-container-toolkit installation"
	@echo ""
	@echo "Testing llama.cpp GPU detection..."
	@docker run --rm --gpus all ghcr.io/ggml-org/llama.cpp:server-cuda /app/llama-server --list-devices 2>&1 || echo "llama-server GPU detection failed"

.DEFAULT_GOAL := up