docs: add compression space verifier
Browse files
README.md
CHANGED
|
@@ -34,6 +34,10 @@ Live Space:
|
|
| 34 |
`python3 scripts/smoke_compression_api.py --base-url https://wchen22-touchdown-compression-classifier.hf.space --include-classify --include-batch`
|
| 35 |
validates `/health`, `/v1/classify`, single `/v1/compress`, and managed
|
| 36 |
`inputs[]` batch.
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
- Latest live result: `/v1/compress` saved 28/70 estimated tokens;
|
| 38 |
managed `inputs[]` returned `input_count=2`, `succeeded=2`, `failed=0`,
|
| 39 |
and `/v1/classify` returned KEEP-only DeBERTa tokenizer labels.
|
|
|
|
| 34 |
`python3 scripts/smoke_compression_api.py --base-url https://wchen22-touchdown-compression-classifier.hf.space --include-classify --include-batch`
|
| 35 |
validates `/health`, `/v1/classify`, single `/v1/compress`, and managed
|
| 36 |
`inputs[]` batch.
|
| 37 |
+
- Full deployment receipt:
|
| 38 |
+
`python3 scripts/verify_compression_space.py --expected-sha <sha> --out reports/generated/compression_space/hf_space_verification.json`
|
| 39 |
+
validates HF runtime metadata, repo/runtime SHA agreement, API smoke, and
|
| 40 |
+
remote/local Space file parity.
|
| 41 |
- Latest live result: `/v1/compress` saved 28/70 estimated tokens;
|
| 42 |
managed `inputs[]` returned `input_count=2`, `succeeded=2`, `failed=0`,
|
| 43 |
and `/v1/classify` returned KEEP-only DeBERTa tokenizer labels.
|