# Model Blob SHA-256 Hashes — Day 4 MCP Tool-Building Lab
# Recorded at M5 build time (2026-07-07) via Ollama registry manifest inspection;
# reused verbatim at M8 (same vendored model — no new models per plan).
#
# Day-4 uses qwen2.5:3b (NOT the 1.5b model from Days 1-3).
# Rationale: 3B class is required for reliable tool-calling (PROJECT.md §9).
# Small models (0.5-1B) emit malformed JSON in tool_call arguments.
#
# These are content-addressed Ollama model blob hashes (ADR-6).
# If the SHA-256 matches after pulling, you have exactly the same model weights.
#
# Verify after pulling:
#   docker compose exec ollama ollama show qwen2.5:3b
#
# Air-gapped verification: check that the blob file exists in the volume:
#   docker compose exec ollama ls -lh /root/.ollama/models/blobs/sha256-<first-32-chars>

# ── qwen2.5:3b ────────────────────────────────────────────────────────────────
# Architecture: qwen2  |  Parameters: 3.1B  |  Quantization: Q4_K_M
# Context length: 32768  |  Embedding length: 2048  |  License: Apache 2.0
# Capabilities: completion + tool-calling (reliable at 3B class — use this for Day-4)
# Tool-calling note: verified to produce well-formed JSON tool_call arguments
#   for 1-2 tool schemas. Performance degrades with >5 tools per call on CPU —
#   relevant once you add your own tools in this lab.
#
# Main GGUF blob (~1.93 GB, Q4_K_M quantized):
QWEN2_5_3B_BLOB_SHA256=5ee4f07cdb9beadbbb293e85803c569b01bd37ed059d2715faa7bb405f31caa6
#
# Config/manifest SHA-256 (from Ollama registry OCI manifest):
QWEN2_5_3B_CONFIG_SHA256=161ddde4c9cd07c9f1ccb4e0167c434bce72caeb3fc1844262fa66bc877b0426

# ── Total download size ───────────────────────────────────────────────────────
# qwen2.5:3b (GGUF Q4_K_M):    ~1.93 GB
#
# Ollama model registry: registry.ollama.ai (not Docker Hub)
# These hashes are for Ollama GGUF model blobs, not Docker image digests.
