Merge pull request #2 from devang007/release/v0.2.0

New model weights. Exact match on the validation set goes from 90.95% → 95.3%, and held-out test / test_ood are measured and published for the first time.

Split	n	Exact match	JSON-valid	Schema-valid
val	4,641	95.3%	99.1%	99.1%
test (held out)	4,641	95.2%	99.0%	99.0%
test_ood	4,775	94.2%	98.8%	98.8%

Retrained on a 2× larger corpus (83,544 train rows, up from 42,558) on a Colab GPU. Latency on 1-thread CPU: p50 532 ms, p95 983 ms per query. Still 15.57M params — the model stays tiny and fully offline.

Upgrading

git pull && ./install.sh

install.sh now pins MODEL_VERSION=v0.2.0, so a fresh install pulls these weights automatically and verifies the sha256.

Assets

• logx-model-v0.2.0.zip (58 MB) — safetensors + JSON only, no pickle files
• logx-model-v0.2.0.zip.sha256 — verified automatically by the installer

Notes

• The tokenizer now adds 50 tokens (the 6 ASCII ones for JSON braces, plus accented-Latin and CJK characters present in the corpus). As always, load the tokenizer shipped in the zip, never the base model's.
• Largest remaining error buckets across all splits: top-K value errors, action confusion, and invalid JSON. Training data is synthetic — test_ood at 94.2% is the honest generalization signal.

Full model card: MODEL_CARD.md

Optimization : cpu only torch dependency

Full Changelog: v0.1.0...v0.1.1

logx v0.1.0

logx v0.1.0 — first release

logx queries nginx logs in plain English, fully offline: a 15.57M-param
fine-tune of google/t5-efficient-tiny translates your question into a
strict JSON DSL, which a deterministic, read-only, shell-free executor runs
over your log files. No network calls, no data egress, Apache-2.0.

git clone https://github.com/devang007/logX && cd logX
./install.sh             # downloads the model asset below, verifies sha256
logx -q "top 5 ips" -src /var/log/nginx/access.log
tail -n 100000 access.log | logx -q "how many 502s in the last hour"
man logx

Model (attached assets)

• logx-model-v0.1.0.zip (56 MB) — weights (safetensors) + tokenizer (JSON).
  No pickle files. See MODEL_CARD.md.
• logx-model-v0.1.0.zip.sha256 — checksum, verified automatically by the installer.

Val metric (n=2,364)	Value
Exact match	90.95%
JSON-valid	98.27%
Schema-valid	98.18%

Scope & safety

• nginx access + error logs (Apache combined access format also parses).
• Out-of-scope questions abstain (exit 2) — nothing executes.
• Execution is restricted to an allowlist of read-only binaries; filter
  values never enter code. Model output is schema-validated before any run.
• Known limits: OOD generalization not yet published; ~4 s model load per
  invocation (ONNX INT8 planned); English only.