Datasets:
|
Download CODE_README.md from shantomathew/agentmart-tooltune-data: direct link, hf CLI and curl.
- Browser
- Download file 4.39 kB
-
https://huggingface.co/datasets/shantomathew/agentmart-tooltune-data/resolve/main/CODE_README.md
- Command line
-
hf download hf://datasets/shantomathew/agentmart-tooltune-data/CODE_README.md
-
curl -L -o CODE_README.md https://huggingface.co/datasets/shantomathew/agentmart-tooltune-data/resolve/main/CODE_README.md
4.39 kB
Lexsi Labs Hackathon — Track 3 (Agentic Fine-tuning): AgentMart Tool-Calling Pipeline
Chains two Lexsi Labs toolkits for reliable tool calling, targeting a Cohere Labs open-weights student model:
- AgentTune — trajectory / post-training data formats (reference clone in
AgentTune_ref/, from https://github.com/lexsi-labs/AgentTune). - AuditKIT (
auditkit==1.1.11from PyPI) — agent evaluation metrics, used through its real API (Sample,ToolCallF1,ToolCallValidity,RedundantToolCalls,TrajectoryMatch,load_agenttune,episodes_from_agenttune). - Student target (GPU step):
CohereLabs/aya-expanse-8b/CohereLabs/tiny-aya-global— seetrain_aya_sft.ipynb.
What is measured locally vs. the GPU step: the executed local run evaluates
two agent policies (naive baseline vs. a policy distilled by code from the
AgentTune trajectories) with AuditKIT scoring. The Cohere fine-tune itself is
the notebook step and has not been run yet — no LLM metrics are claimed.
See ARTICLE.md for the full write-up and the real before/after table.
Layout
| Path | What |
|---|---|
src/tasks.py |
5 tool schemas + deterministic tool implementations + 170-task suite (130 train / 40 held-out test, entity/value-disjoint) |
src/generate_data.py |
Builds data/tasks.json, data/sft_agenttune.jsonl, data/trajectories.jsonl |
src/policies.py |
BaselinePolicy (naive keyword agent) + LearnedPolicy (naive-Bayes router + learned cue/shape extractor, fitted on the SFT file) |
src/evaluate.py |
Harness: scores both policies on the test split (own metrics + AuditKIT), verifies AgentTune→AuditKIT parsing, writes results/metrics.json |
data/sft_agenttune.jsonl |
130 SFT rows in AgentTune EventLog.as_dataset_rows('sft') format (messages, segment_weights, loss_mask) |
data/trajectories.jsonl |
130 TrajectoryDataset records (AgentTune trainer format; AuditKIT-readable) |
results/metrics.json |
Real before/after numbers from the executed run |
RUN_OUTPUT.txt |
Full console output of the executed pipeline |
train_aya_sft.ipynb |
GPU notebook: TRL SFT (LoRA) on Aya Expanse 8B + AgentTune-native alternative + AuditKIT hand-off (not executed locally) |
ARTICLE.md |
HF-Article-style write-up with the real metrics |
AgentTune_ref/ |
Shallow clone of Lexsi-Labs/AgentTune used as the format source (README "Hand the run to AuditKIT"; src/agenttune/agentic/events.py for the SFT row schema) |
Reproduce (Python 3.12, CPU only)
cd lexsi-agentmart
python3 -m venv .venv
.venv/bin/pip install auditkit==1.1.11 # or: .venv/bin/pip install -r requirements.txt
.venv/bin/python src/generate_data.py
.venv/bin/python src/evaluate.py # prints the report, writes results/metrics.json
Expected headline result (held-out test, n=40): task success 0.15 → 1.00,
tool-selection accuracy 0.75 → 1.00, AuditKIT redundant calls
0.1375 → 0.00 — with the in-distribution caveat documented in ARTICLE.md.
Formats — where each came from
- SFT rows: AgentTune
EventLog.as_dataset_rows(fmt='sft')(AgentTune_ref/src/agenttune/agentic/events.py): one{"messages", "segment_weights", "loss_mask"}row per trajectory; tool calls rendered as<tool_call>{"name": ..., "arguments": {...}}</tool_call>assistant messages (weight 2.0), tool results astoolmessages (weight 0.0). - Trajectories: AgentTune README, "Hand the run to AuditKIT": one JSONL
record per rollout —
{"task", "steps": [{"action": {"tool_calls": [{"id", "name", "arguments"}]}, "observation"}], "final_response", "metadata": {"conversation": [...]}}. Verified:auditkit.load_agenttune()returns 130 samples andepisodes_from_agenttune()130 episodes for our file. - AuditKIT scoring:
Sample(input, tools=<OpenAI-format schemas>, expected_tool_calls=[[gold]])scored withmetric.score(sample, output, {"trace": {"tool_calls": turns}}).
Notes
- Train/test split: prompts fully disjoint; entities and values (topics,
expressions, people, symbols, cities, dates) disjoint across splits;
template wording families shared — disclosed in
ARTICLE.mdLimitations. - Nothing in this repo has been pushed to GitHub or Hugging Face, and no submission form has been filled — that step belongs to the team lead.