beyoru commited on
Commit
1e7143f
·
verified ·
1 Parent(s): 4f373f2

Note: text input only

Browse files
Files changed (1) hide show
  1. README.md +3 -0
README.md CHANGED
@@ -30,6 +30,8 @@ You give gemmaExtract a document and a list of fields you want. It returns each
30
  document ──► one forward pass ──► {field: value, span, p_null, confidence}
31
  ```
32
 
 
 
33
  ## Quick start
34
 
35
  ```python
@@ -132,6 +134,7 @@ Measured on half of an H200, bf16, batch 1, default null threshold.
132
  - **Span-based**: answers that must be paraphrased or merged from scattered places (long free-form answers, multi-line addresses) are weaker than with a generative model.
133
  - Up to **2,048 tokens per pass**; longer documents are truncated (`usage.truncated` tells you).
134
  - Languages other than English and Vietnamese work (see the XQuAD table) but absent-field detection has only been checked in English and Vietnamese.
 
135
  - Not supported yet: `time`, `duration`, `estimate`, `enum` fields and image input.
136
  - Runs with `transformers` (≥ 5.17); there is no serving-engine integration yet.
137
 
 
30
  document ──► one forward pass ──► {field: value, span, p_null, confidence}
31
  ```
32
 
33
+ > **Text input only.** gemmaExtract currently reads plain text (`state` must be a string). For scanned documents, photos or PDFs, run OCR first and pass the recognized text. Image input is not supported yet.
34
+
35
  ## Quick start
36
 
37
  ```python
 
134
  - **Span-based**: answers that must be paraphrased or merged from scattered places (long free-form answers, multi-line addresses) are weaker than with a generative model.
135
  - Up to **2,048 tokens per pass**; longer documents are truncated (`usage.truncated` tells you).
136
  - Languages other than English and Vietnamese work (see the XQuAD table) but absent-field detection has only been checked in English and Vietnamese.
137
+ - Text input only: images, scans and PDFs need an OCR step first.
138
  - Not supported yet: `time`, `duration`, `estimate`, `enum` fields and image input.
139
  - Runs with `transformers` (≥ 5.17); there is no serving-engine integration yet.
140