Downloads · 30 days
16
41% of all-time downloads
ririye/restaurant-finder-qwen25-7b-lora
restaurant-finder-qwen25-7b-lora is a text generation model from ririye. Use it when you need the model to write or continue text. It is set up for peft. The card lists the license as apache-2.0.
A LoRA adapter for Qwen/Qwen2.5-7B-Instruct that teaches the base model to drive the anything-finder restaurant-search agent: a four-tool ReAct loop over OpenStreetMap data (Nominatim geocoding + Overpass POI search).
Downloads · 30 days
16
41% of all-time downloads
All-time downloads
39
Public
Repo size
334 MB
Likes
0
Public
Click a slice to open those files.
.safetensors162 MB · 91%
From the Hugging Face model README
A LoRA adapter for Qwen/Qwen2.5-7B-Instruct that teaches the base model to drive the
anything-finder restaurant-search agent: a four-tool ReAct loop over
OpenStreetMap data (Nominatim geocoding + Overpass POI search).
The adapter is distilled from Claude Sonnet trajectories — the teacher ran the real agent against a local Dallas OSM extract, and every tool call, tool result, and final recommendation was captured verbatim.
| Tool | Purpose |
|---|---|
read_food_preferences | Read the user's stored preference profile. |
get_current_location | Fall back to the configured home city/state. |
geocode_location | Place phrase -> coordinates (Nominatim). |
search_restaurants | Nearby eateries within radius_m (Overpass); widen and retry when sparse. |
Tool JSON schemas ship with this repo as tool_schemas.json, generated from the
live @tool objects so they cannot drift from the app.
Records are OpenAI-style tool-calling chat transcripts built from the agent's
full message log — not flattened prompt/completion pairs. See
src/training/dataset.py in the source repo for the validation rules (a row is
dropped when a tool_call_id is unresolved or unanswered, when the final turn
is not a tool-call-free assistant message, or when that turn is empty).
Qwen/Qwen2.5-7B-Instruct (bfloat16)16 / 32 / 0.05q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj4096 (overlong rows: drop)3, LR 0.0001 (cosine), optimizer adamw_torch_fused4 x 2 grad-accum202609020.41840.5309{% generation %}, so TRL's assistant_only_loss cannot be used. Labels are
masked by rendering messages[:i] and messages[:i+1] and unmasking the
delta. System, user, and tool spans contribute no loss; every assistant turn
does.enable_thinking: false — the captured trajectories carry no
reasoning content, so prompt it the same way at inference.vllm serve Qwen/Qwen2.5-7B-Instruct \
--enable-lora --lora-modules af-lora=/adapters/af-lora \
--max-lora-rank 16 --api-key EMPTY
Then point the app at it:
LLM_BACKEND=lora LLM_MODEL_LORA=af-lora LLM_BASE_URL=http://localhost:8000/v1
The source repo renders a three-way comparison (Claude Sonnet vs. the raw base
model vs. this adapter) over an identical slice of queries — final answers plus
telemetry (tool paths, per-call tokens, latency, cost) — with make compare.